A reading practice · Greek from the 1st century CE, Hebrew from the 2nd; the English ciphers modern

Gematria

letters as numbers in Hebrew, Greek, Arabic and Latin

Your name, or any word, read as a number: the sum of its letters under the ciphers of Hebrew, Greek, Arabic and Latin.

Draw your free chart No email needed

Get the reading · $4.99

Your Gematria reading, $4.99. Includes this system's raw data. Written for you by AI. Read a sample. Charts and the written text are free in every account.

Your name's gematria is a number: on the English Ordinal cipher, where A is 1 and Z is 26, the sample name Jason adds up to 59, and the chart gives every word of a name its total under 24 ciphers from four scripts, each labelled by where it comes from.

The Greeks called the practice isopsephy and the rabbis gematria, Arabic writers counted by the abjad, Agrippa gave the Latin letters values in 1533, and the English ciphers were named online in the 2000s and 2010s.

At a glance

What it is: Adding up the number values of a word's letters: gematria in Hebrew, isopsephy in Greek, the abjad in Arabic
Attested from: Greek isopsephy in the 1st century CE; gematria in rabbinic teaching from the 2nd
The letters as numerals: Greek by the 5th century BCE, Hebrew by 78 BCE, Arabic before the 8th century CE
The Latin table: Agrippa, De occulta philosophia, Book II, chapter xx (Cologne, 1533)
The English ciphers: Modern: named on the calculator sites of the 2000s and 2010s
What you enter: Any text; a Hebrew, Greek or Arabic value needs the text written in that script
Our ciphers: 24: thirteen Latin, seven Hebrew, two Greek, two Arabic; 23 have a page
Evidence: A device for reading texts, which the Talmud calls a support and not a proof; with many ciphers, matches are certain

Sample name

Jason Carlos Eusebio · 18 letters, Latin script

J 10
A 1
S 19
O 15
N 14

59 in all

C 3
A 1
R 18
L 12
O 15
S 19

68 in all

E 5
U 21
S 19
E 5
B 2
I 9
O 15

76 in all

The English Ordinal, A 1 to Z 26: 59 + 68 + 76 = 203.

English Ordinal: 203 203 → 5 A = 1 … Z = 26; the alphabet position; the "Simple" gematria of the popular sites
Full Reduction: 68 68 → 14 → 5 A = 1 … I = 9, then 1 again; the Pythagorean numerology table
Reverse Ordinal: 283 283 → 13 → 4 Z = 1 … A = 26; a modern mirror of the ordinal
Reverse Reduction: 112 112 → 4 the reverse ordinal reduced to one digit; a modern mirror of the reduction
Latin (Agrippa): 1,386 1,386 → 18 → 9 A 1 … I 9, K 10 … S 90, T 100, U 200, X 300, Y 400, Z 500, J 600, V 700, W 900; Agrippa, 1533, with J, U, V and W placed as the popular sites place them; a Latin table, not a Hebrew one
Ordinal x 6 (Sumerian): 1,218 1,218 → 12 → 3 the ordinal times six, A = 6 … Z = 156; a modern cipher; the name is a borrowing, not a Sumerian source

Behind “More ciphers”

Chaldean 65
Standard 986
Offset 36 833
Septenary 64
Trigonal 1,656
Squares 3,109
Primes 642

No Hebrew, Greek or Arabic value: the name is written in Latin letters, and the chart does not transliterate it.

Figure 1. The gematria of a fictional name, Jason Carlos Eusebio, as the chart gives it. Each word is added up letter by letter on the English Ordinal (A 1 to Z 26); below, the totals under the six core Latin ciphers with the digit sums that reduce each to one figure and the line that says where each cipher comes from, then the seven ciphers behind "more". None of its totals lands on a famous value; when one does, the chart names the word, and it reports a match across scripts as a coincidence of numbers, not as a correspondence the tradition draws. Method: the name upper-cased with accents stripped; each Latin letter valued on each cipher's table and the words added; every total reduced to its digital root, the Chaldean total read as Cheiro's compound. A name in Latin letters is not transliterated, so it has no Hebrew, Greek or Arabic value.

On this page

How it works
Origins and history
The ciphers
Variants and what we leave out
Myths and disputed claims
Stories and lore
Evidence and criticism
Terms
Sample reading
Sources

How it works

Before separate digits came into use, the Greeks, the Jews and the Arabs wrote numbers with the letters of their alphabets. The first nine letters stood for 1 to 9, the next nine for 10 to 90 and the rest for the hundreds, so any written word could also be read as a sum. Gematria is that reading: add the values of a word's letters and set the word beside others with the same total.

Script	Units	Tens	Hundreds	Our methods
Greek	α to θ, with ϛ for 6	ι to ϟ	ρ to ϡ	Isopsephy; the ordinal behind "more"
Hebrew	א to ט	י to צ	ק to ת (400)	Seven methods, from the standard value to two letter substitutions
Arabic	ا to ط	ي to ص	ق to غ (1000)	The eastern and the western abjad
Latin	A to I	K to S	T to Z, then J, U, V, W	Agrippa's table of 1533 and twelve English ciphers

The scripts are kept apart. Text in Latin letters is scored under the Latin ciphers only. A value in Hebrew, Greek or Arabic needs the text written in that script, as its speakers write it: we do not transliterate a name from one alphabet into another, because a letter-for-letter swap gives spellings nobody writes and values no two tools agree on. A text that mixes scripts is scored script by script, and the chart says so.

Before counting. The text is upper-cased and stripped of accents, Hebrew vowel points and Arabic short-vowel marks. Letters such as Ø, Æ and ß are folded to O, AE and SS. In Hebrew a final letter counts as its ordinary form, except in Mispar Gadol; in Greek the final sigma counts as sigma; in Arabic the forms of alif with hamza count as alif and the final ta' marbuta as ha'.

Each total, three ways. For every cipher the chart gives the total, its digital root (the digits added until one is left) with the chain of sums, and the letters with their values word by word. Cheiro's Chaldean table is read as he read it, a compound number with its single digit, exactly as on our Numerology chart.

The ciphers differ a great deal. The same word gives very different numbers depending on the table. GEMATRIA itself, for instance, under each of our thirteen Latin ciphers:

Cipher	Rule	GEMATRIA
English Ordinal	A 1 to Z 26	74
Full Reduction	A to I 1 to 9, then again	38
Reverse Ordinal	Z 1 to A 26	142
Reverse Reduction	the reverse ordinal reduced	52
Latin (Agrippa)	units, tens and hundreds, Agrippa 1533	233
Ordinal x 6 (Sumerian)	the ordinal times six	444
Chaldean	Cheiro's 1 to 8	21
Standard	units, tens and hundreds in alphabet order	353
Offset 36	A 36 to Z 61	354
Septenary	1 to 7 and back	32
Trigonal	the triangular number of the ordinal	562
Squares	the ordinal squared	1,050
Primes	the 26 first primes	228

In Hebrew the word שלום, shalom, "peace", is 376 by the standard method, 936 when its final mem counts 600, 52 by the letters' places, 16 with the zeros dropped and 7 as a digital root; exchanged letter for letter by Atbash it gives 112, by Albam 93.

Matches. The point of the practice is the equivalence: two words with the same value, which the tradition reads together. The chart keeps a short table of famous values of ordinary words from the Hebrew and Greek traditions, each recomputed in our tests: in Hebrew love and one (13), life (18), wisdom (73), king (90), light and secret (207), serpent (358), peace (376), truth (441), the two Nero Caesar readings (616 and 666) and the first word of Genesis (913); in Greek love (93), holy and good (284), the Nile (365) and the dove (801). No sacred name is among them: the chart never sets a name's total beside a name or title of God, a prophet, a saint, a messiah or a deity, a scripture or a sacred formula. When one of your totals lands on a famous value, the chart names it and says whether the script and the method agree. A match across scripts, an English total that equals a Hebrew word, is a coincidence of numbers that no tradition draws, and the chart says that too.

What we do not do. No number here is read as a message. We do not read events, dates or news by number, and we do not compute letter skips in scripture.

Origins and history

The alphabets were numerals for centuries before anyone is known to have added up a word; the practice, once it appears, belongs to three traditions, and the English ciphers are the newest layer of all.

When	Who	What happened
717 to 706 BCE	Sargon II of Assyria	His Cylinder Inscription says the wall of his new city was made "16,283 cubits, the numeral of my name" (Luckenbill's translation). Cuneiform has no letter values, so how the number was reached is unknown.
By the 5th century BCE	Ionian Greeks, around Miletus	The alphabetic numerals: 1 to 9, 10 to 90 and 100 to 900 on the letters, with three old letters kept for 6, 90 and 900. In official use in Greece by about 200 BCE.
3rd century BCE	Apollonius of Perga	The earliest known number play on a Greek verse, by multiplying the letters' values.
About 200 to 78 BCE	Judean coins and scribes	The Hebrew letter numerals, modelled on the Greek, appear; the earliest dated coins carry them.
Before 79 CE	Pompeii; Leonidas of Alexandria	Graffiti such as "I love her whose number is 545"; Leonidas writes epigrams whose couplets add up to the same total.
After 59 CE	Roman wits, recorded by Suetonius	Nero's name and the words "killed his own mother" both make 1005.
70 to 132 CE	The Epistle of Barnabas	Reads the 318 servants of Abraham (Genesis 14:14) as the letters of Jesus' name and the cross.
2nd century CE	Rabbi Eleazar Hisma; the thirty-two rules	The Mishnah names gematria (Avot 3:18); the rules of interpretation attributed to Rabbi Eliezer ben Jose the Galilean list it.
About 180 CE	Irenaeus of Lyon	Attacks Gnostics who read Jesus as 888, and discusses 666 and the variant 616 without naming anyone.
Late 3rd or early 4th century	P115, a papyrus from Oxyrhynchus	The oldest surviving manuscript with 616, published in 1999.
About 200 to 500 CE	The Talmud and the midrash	Gematria used as a support (asmakhta), not a proof: wine and secret are both 70.
7th century onward	Arabic scribes	The abjad: the 28 letters in the old Semitic order as numerals up to 1000, in an eastern and a western order; kept for chronograms after positional digits arrived.
12th and 13th centuries	The Hasidei Ashkenaz; Eleazar of Worms	Gematria applied on a large scale in commentary and prayer; Nahmanides later tries to limit it.
1205	Judah al-Harizi	The earliest known Hebrew chronogram, in a poem; dates in epitaphs follow from the 13th century.
1240 to after 1291	Abraham Abulafia	The "Kabbalah of names": letter combination and permutation as a contemplative practice.
1340 to 1394	Fazlallah Astarabadi	Founds Hurufism, a Sufi letter mysticism of the Arabic and Persian alphabets.
1389 or 1390	The tomb of Hafez, Shiraz	A Persian chronogram gives the poet's death year, 791 of the Hijra, in the letters of a phrase.
1523	John Skelton	Hides a name in two rows of numbers, one per letter: a substitution cipher, not gematria.
1533	Heinrich Cornelius Agrippa	De occulta philosophia, Book II: the Greek, Hebrew and Latin letter tables side by side; the Latin one is new.
1831 to 1837	Fritzsche, Benary, Hitzig, Reuss	Four scholars independently read 666 as Nero Caesar in Hebrew letters.
1865 to 1869	Leo Tolstoy	In War and Peace Pierre Bezukhov makes Napoleon 666 on a French table.
1887	S. L. MacGregor Mathers	A table of Hebrew letters with Roman equivalents, the route by which occultists gave English words Hebrew values.
1906	Schechter and Levias	The Jewish Encyclopedia's article, still the standard short account of the methods.
1920s	Cheiro	Prints the "Chaldean" letter table, 1 to 8.
1976	James Lees	English Qaballa, the one modern English cipher with a documented origin.
1994 and 1999	Witztum, Rips and Rosenberg; McKay and colleagues	The "Bible code" paper and its refutation, both in Statistical Science.
2007	P. J. Williams	Shows that P115's 616 does not settle the Nero reading.
By February 2010	gematrix.org	The most used English calculator, with the ordinal, the ordinal times six and Agrippa's table, and a phrase list fed by its users.
2013 onward	Video channels	Reading the news by gematria on YouTube, which made the practice notorious.
By January 2017	Gematrinator (Derek Tikkuri)	A calculator of dozens of ciphers; renames Agrippa's table "Latin".
2026	EsoTerrier	The engine rebuilt: Agrippa's table added, the mirror ciphers dropped, the Hebrew methods extended, transliteration removed, a checked table of famous values.

Greek. The alphabetic numerals were in place by the 5th century BCE, but the summing of words is attested only in the Roman period. A wall in Pompeii carries a lover's "whose number is 545"; Leonidas of Alexandria, a Greek poet in 1st-century Rome, wrote couplets of equal sums; and after Nero had his mother killed, Suetonius records the verse that went round Rome, "a calculation new: Nero his mother slew", because Νέρων and the Greek for "killed his own mother" both make 1005. Christians took up the device: the Epistle of Barnabas finds Jesus and the cross in Abraham's 318 servants, and Irenaeus, about 180 CE, has to argue with Gnostics who built doctrine on the 888 of Jesus' name.

A page of a seventeenth-century English book headed Chap. XVIII, Of the notes or figures of the Græcians. A short paragraph, then a first table numbering the Greek letters in order from 1 to 24, a second paragraph, and a second table in three rows giving the letters as units, tens and hundreds, with the three extra numeral signs for 6, 90 and 900. — Figure 2. Agrippa, *Three Books of Occult Philosophy*, in the English of 1651, Book II, chapter xviii (first printed in Latin in 1533): "Of the notes or figures of the Græcians". The first table counts the twenty-four letters in order, 1 to 24; the second gives them as units, tens and hundreds, with the three numeral signs for 6, 90 and 900 that the alphabet had otherwise dropped. The second order is isopsephy's. 1651. Wellcome Collection, Public Domain Mark 1.0.

Hebrew. The Hebrew letter numerals were adapted from the Greek in the last two centuries BCE, and the word gematria, itself probably from the Greek, appears in the Mishnah in the 2nd century CE. There Rabbi Eleazar Hisma calls the reckoning of the equinoxes and gematria the parperaot of wisdom, its side dishes rather than its meal; the thirty-two rules of interpretation list gematria among the ways of reading a verse. The rabbis' classic case is Abraham's 318 servants read as Eliezer, his one servant, whose name makes 318. The Talmud uses such readings as an asmakhta, a support for a teaching already held, and later authorities such as Ibn Ezra and Nahmanides held that it could support a teaching but not establish one; Nahmanides accepted only the gematrias received from tradition. The Hasidei Ashkenaz of the 12th and 13th centuries used gematria far more widely, in commentary on law and liturgy, and in the Kabbalah it became one of the main tools for reading scripture.

A detail of a seventeenth-century English book page under the line "Now the Classes of the Hebrew numbers are these": three rows of Hebrew letters, read from right to left, under the numbers 1 to 9, 10 to 90 and 100 to 900, the last row running on from tav at 400 to the five final letter forms at 500 to 900. — Figure 3. The next chapter of the same book, xix, "Of the notes of the Hebrews", page 232: the letters in three classes, units, tens and hundreds, read from right to left. The hundreds run on past tav (400) with the five final forms for 500 to 900, the large count that Mispar Gadol keeps. A Christian Renaissance author copying the Jewish numeral letters into a book of learned magic; the letters are shown as numbers only. 1651. Wellcome Collection, Public Domain Mark 1.0.

Arabic. The abjad keeps the old Semitic order of the letters, abjad hawwaz hutti, and adds the six letters peculiar to Arabic for 500 to 1000. Positional digits took over arithmetic from the 8th century, but the letter values stayed in use for numbering lists and for chronograms, phrases whose letters give the date of what they record, such as the line on Hafez's tomb. Hurufism, from the 14th century, made the letters themselves the centre of a mystical teaching. In South Asia the number 786, the basmala's value in the eastern order, is written on shop signs, letterheads and the top of letters in place of the words themselves: a South Asian custom, which many scholars and most of the Arab world do not follow.

Latin. The Latin alphabet was never a numeral system in this way. In 1533 Agrippa, after giving the Greek and Hebrew tables, wrote that "the Pythagorians say ... that the very Elements of letters have some certain divine numbers" and set out Latin values on the Hebrew pattern, using consonantal I and V, HI and HV to fill the twenty-seven places, together with a rule for finding a person's ruling planet from a name's number divided by nine. We found no recorded use of his table between his day and the calculator sites; Ashe argues that he never used it himself.

The top of a seventeenth-century English book page, Of Occult Philosophy, Book II, page 236. Two rows of Latin capital letters stand under their numbers: A to I as 1 to 9, K to N as 10 to 40, then O to S as 50 to 90, T, V, X, Y and Z as 100 to 500, and I, V, HI and HV as 600 to 900. — Figure 4. Chapter xx, page 236: Agrippa's Latin letters on the same pattern, A to I 1 to 9, K to S 10 to 90, T to Z 100 to 500, then consonantal I and V, HI and HV for 600 to 900. It is the first known table of values for the Latin alphabet. The calculator sites fit it to the 26-letter alphabet by moving J, U, V and W, and our Latin (Agrippa) cipher follows them. 1651. Wellcome Collection, Public Domain Mark 1.0.

The English ciphers. The Golden Dawn and its heirs gave English words Hebrew values through Mathers' table of equivalents, and Cheiro's "Chaldean" table of the 1920s was built from Hebrew values reduced to one digit. The ciphers most calculators show first are younger still. When the Wayback Machine first saw gematrix.org in February 2010 it already offered the ordinal, the ordinal times six and Agrippa's table under a religious label it has nothing to do with; Gematrinator, a website by 2017, lists dozens more and renamed Agrippa's table "Latin". From 2013 video channels used these calculators to read the news, and the practice became a byword for conspiracy theories. Who first named most of the community ciphers is not known.

The ciphers

Each cipher the chart computes has its own page, except the Greek ordinal, which the chart offers behind "more" for comparison: the rule and the full letter table, where the cipher comes from and how sure that is, the names other sites give it, worked examples, and the claims made about it.

Latin script: the six core ciphers

Latin script: seven more

Hebrew script

Greek script

Isopsephy

Arabic script

The Latin core six are the ciphers the popular calculators open with, plus Agrippa's; the other seven are shown behind "more ciphers". The Hebrew, Greek and Arabic methods apply only to text written in those scripts.

Variants and what we leave out

Five mirror ciphers. Reverse versions of the Ordinal x 6 (Sumerian), Offset 36, Trigonal, Squares and Primes ciphers, which our old chart showed as equals: dropped. Reverse Ordinal and Reverse Reduction stay, because every calculator opens with them.
Transliteration. Giving an English name Hebrew, Greek or Arabic values by swapping letters, as some sites do with Mathers' table: not offered.
Agrippa's own 27 values. His table doubles I and V as consonants and has HI 800 and HV 900. No strict Agrippa cipher is offered; the chart uses the 26-letter placement every site uses, labelled as the popular sites' placement of J, U, V and W.
The "Offset 36" name. The sites call that cipher "Satanic", explaining that the 36th triangular number is 666. We use a name that says what it does.
Other English ciphers. English Qaballa, the Bacon ciphers, keypad and Fibonacci ciphers and the rest of Gematrinator's list: not computed.
More Hebrew methods. TorahCalc and Kesher offer some twenty-five, among them the cumulative, squared and spelled-out values: not yet computed, so no page calls our seven complete.
Variant counts. Some tools give the Hebrew finals 23 to 27 in the ordinal, write Greek stigma as sigma and tau, or ignore a lone hamza; ours counts finals as their base, stigma as 6 and hamza as 1.
Word lists. "Words with the same value" lookups over the Hebrew Bible and the Greek New Testament from open texts are later work; no calculator's user-fed phrase list is used.
Letter skips. The so-called Bible code is not gematria and is not computed.

Myths and disputed claims

What is said about gematria, who says it, and what the record shows. The label is ours.

Isopsephy, and with it gematria, began with Pythagoras and his school at Miletus in the 6th century BCE.
disputed

Who says it: Popular histories and our retired research file.

What the record shows: The alphabetic numerals arose around Miletus and are attested by the 5th century BCE, but no early source has Pythagoras number letters, and the summing of words is attested only from the 1st century CE.
Sargon II made the wall of his new city 16,283 cubits long to match the number of his name, the first gematria.
unverified

Who says it: Encyclopaedias, after the inscription.

What the record shows: The inscription does say "the numeral of my name" (Luckenbill's translation). How the scribes reached the figure cannot be checked: cuneiform is not an alphabet and has no letter values.
Hebrew letters stood for numbers in the time of Moses and the prophets, so the Torah was composed with gematria in mind.
false

Who says it: Many religious sites.

What the record shows: The Hebrew letter numerals were adapted from the Greek between about 200 and 78 BCE, the date of the earliest coins with them. A text written before they existed cannot have been composed with them.
The Kabbalists were the first to use gematria widely.
disputed

Who says it: Our retired research file.

What the record shows: The rabbis of the 2nd century CE come first; the Hasidei Ashkenaz, Eleazar of Worms and Abulafia expanded the practice greatly in the Middle Ages.
The word gematria comes from the Greek geometria.
disputed

Who says it: Most dictionaries.

What the record shows: Possibly; others derive it from grammateia, and the Jewish Encyclopedia already recorded the dispute in 1906.
666 is proven to mean Nero Caesar, and 616 proves it.
disputed

Who says it: Many commentaries and gematria sites, and our own earlier page.

What the record shows: Nero Caesar in Hebrew letters makes 666, and the Latin spelling 616: the usual scholarly reading since the 1830s. But Irenaeus, about 180 CE, knew no such reading, and the Greek word for "beast" written in Hebrew letters gives the same two numbers. A reading, not a proof.
The Gospel writers meant the name Jesus to make 888, a divine signature.
unverified

Who says it: Christian numerology pages; our retired research file.

What the record shows: The value is right and ancient: the Sibylline Oracles set it as a riddle, and the Gnostics whom Irenaeus attacked used it. That the evangelists intended it is claimed by none of the early sources and cannot be shown.
The name Abraxas was made so that its letters would give 365, the days of the year.
told as lore

Who says it: Irenaeus, reporting on Basilides; gnostic tradition.

What the record shows: The sum is right, and Irenaeus links the name to 365 heavens. That the name was built for the number is the ancient account, told as such.
786 is sacred throughout Islam and may stand for the basmala anywhere.
disputed

Who says it: Many sites.

What the record shows: 786 is the basmala's value in the eastern order (1,026 in the western). Writing 786 for it is a South Asian custom, which many scholars and most of the Arab world do not follow.
The cipher A 1 to I 9, K 10 to S 90, T 100 and on is a Jewish or Hebrew method.
false

Who says it: gematrix.org, the sites that copy it, and videos built on it.

What the record shows: It is Agrippa's Latin table of 1533 fitted to the 26-letter alphabet. No Jewish tradition uses it. Gematrinator renamed it "Latin" for that reason, and Ashe documents how the old label feeds antisemitic conspiracy talk. We call it Latin (Agrippa).
Agrippa's letter table dates from 1531.
false

Who says it: Our retired research file.

What the record shows: Book I of his work was printed in 1531; the letter table is in Book II, first printed in the complete Cologne edition of 1533.
The English ciphers with names such as Septenary, Trigonal and Ordinal x 6 (Sumerian) go back to antiquity or to some source outside the modern gematria sites.
false

Who says it: Sites that list them without saying where they come from.

What the record shows: They were named by the online English-gematria community of the 2000s and 2010s. The one modern English cipher with a documented origin, English Qaballa (1976), we do not compute.
The Ordinal x 6 (Sumerian) cipher comes from Sumer or from its counting in sixties.
false

Who says it: Calculator sites.

What the record shows: It is the ordinal multiplied by six. Sumerian cuneiform has no alphabet to value; the name is a borrowing.
The Chaldean cipher is Babylonian.
false

Who says it: Chaldean numerology sites; our earlier page.

What the record shows: It is Cheiro's table, printed in the 1920s. Its values are, for 23 of 26 letters, the Hebrew letter values reduced to one digit, as Sepharial had printed them in 1911.
Mispar Katan is the digital root of a word's value.
false

Who says it: Our earlier page.

What the record shows: Mispar Katan drops the zeros of each letter before adding (yod 1, qof 1, tav 4). The digital root of the standard total is Mispar Katan Mispari. The chart computes both, correctly named.
A name in Latin letters can be given a Hebrew, Greek or Arabic value by swapping its letters one for one.
false

Who says it: Our engine before October 2026; many calculators.

What the record shows: A letter-for-letter swap gives spellings no speaker writes and values no tool agrees on. A value in another script needs the name written in it, as its speakers write it.
Greek architects used isopsephy for the proportions of buildings.
unverified

Who says it: Our retired research file; some sites.

What the record shows: No source was found. An architect of the 2nd century CE, Galen's father, wrote isopsephic verse; that is not evidence that buildings encode names.
John Skelton used Latin gematria in 1523 to hide a name in his Garland of Laurel.
false

Who says it: Our retired research file.

What the record shows: He did hide a name in numbers, but each letter became a number and nothing was added: a substitution cipher, not gematria.
"L'Empereur Napoléon" makes 666.
told as lore

Who says it: Tolstoy's Pierre Bezukhov, in War and Peace.

What the record shows: A novelist's scene. Pierre gets 666 only by bending the spelling, which is part of Tolstoy's joke.
Words or names with the same value are connected, so a match shows a link between them.
false

Who says it: Gematria sites and channels; our page before October 2026.

What the record shows: Reading equal values together is the tradition's way of interpreting a text, a support and not a proof in the Talmud's own terms. With dozens of ciphers and millions of phrases, every total matches something.
Letters found at equal skips in the Hebrew Bible prove that it foretells events.
false

Who says it: Witztum, Rips and Rosenberg (1994); popular books since 1997.

What the record shows: Refuted in the same journal in 1999: the result depended on choices made in gathering the data and disappeared in new tests. Letter skips are not gematria, and we do not compute them.
Gematria shows that news events are scripted by their numbers.
false

Who says it: Videos and forums built on the calculator sites since 2013.

What the record shows: A search with unlimited freedom always finds a match; fact-checkers have shown it. Nothing on this site reads events, dates or news by number.
Our engine finds matching values across Hebrew, Greek, Arabic and Latin at once.
false

Who says it: Our public page before October 2026.

What the record shows: The old engine found no matches at all. The rebuilt one checks totals against a short table of famous values and says when a match crosses scripts.

Stories and lore

These are told as their sources tell them; some are history, some legend, and each says which.

The king's wall. Sargon II had it written that the wall of his new capital measured "16,283 cubits, the numeral of my name". No one now knows how his scribes made the sum, and the line stands alone in cuneiform as a boast that a name is a number.

The number of her honourable name. On a wall in Pompeii someone wrote that he loved a girl "whose number is 545", and another that "the number of her honorable name is 45": isopsephy as flirtation, buried in 79 CE.

Nero's lampoon. Suetonius records the line that went round Rome after the death of Agrippina: "a calculation new: Nero his mother slew", since the emperor's name and the phrase add up alike.

Abraham's 318. Genesis says Abraham armed 318 servants. The Epistle of Barnabas reads the number as IH, the first letters of Jesus' name (18), and T, the cross (300); the rabbis read it as the name Eliezer, his one servant, whose letters make 318. Two traditions used the same device on the same verse to opposite ends.

Wine in, secret out. The Talmud notes that יין, wine, and סוד, secret, are both 70: when wine goes in, the secret comes out.

Love and one. אהבה, love, and אחד, one, are both 13: a staple of Jewish teaching.

The beast's number. Irenaeus, about 180 CE, already had to argue that 666 was the right reading "in all the most approved and ancient copies" and that 616 came from a copyist's slip; he tried several Greek names and declined to choose. Seventeen centuries later four scholars found Nero Caesar in Hebrew letters, נרון קסר, 666, with the Latin spelling נרו קסר giving 616. In 1999 a papyrus with 616 made the news. It is the usual scholarly reading, and still a reading: the word for "beast" written in Hebrew letters gives both numbers too.

Jesus 888. The Sibylline Oracles set the name as a riddle, "eight ones, and as many tens on these, and yet eight hundred" (Terry's 1899 translation), and the Gnostics whom Irenaeus attacked built on it. The sum is ancient; that the Gospel writers meant it is nobody's claim in the early sources.

Abraxas and the year. Basilides, as Irenaeus reports him, taught 365 heavens, and the name Abraxas adds up to 365, the days of the year.

Skelton's adversary. In 1523 John Skelton ended a Latin satire in his Garlande of Laurell with two rows of numbers instead of a name. Numbering the vowels 1 to 5 and the consonants by their place in the Latin alphabet turns them into ROGERUS STATHUM; that key is our reading, since the edition gives none. Each letter becomes a number and nothing is added, so it is a cipher, not gematria.

Pierre's calculation. In War and Peace Pierre Bezukhov, using a French table that values the letters Hebrew-fashion, makes "L'Empereur Napoléon" 666, then forces his own name to the same total and decides he is destined to stop the emperor.

Evidence and criticism

What is documented. The alphabetic numerals and their dates are epigraphic fact, and the uses of isopsephy and gematria in the sources above are textual fact. So is the rabbis' own view of the device: the Talmud treats a gematria as support for a teaching, not as its proof, and medieval authorities said so explicitly.

Equal values are certain. Any table that maps many words onto a few thousand numbers makes coincidences unavoidable; the mathematician Underwood Dudley calls this the law of small numbers. With dozens of ciphers and a list of millions of phrases, every total matches something, which Lead Stories demonstrated in 2022 for the videos that read the news by gematria. A match is a fact about arithmetic.

The Bible code. In 1994 Statistical Science published a paper claiming that names and dates were encoded at equal letter skips in Genesis. In 1999 the same journal published the reply of McKay, Bar-Natan, Bar-Hillel and Kalai, who showed that the result depended on choices made in gathering the data and found nothing in fresh tests.

666 and 616. Modern critical scholarship broadly reads the number as Nero Caesar, but P. J. Williams (2007) shows that the oldest manuscript with 616 does not settle it, since the Greek word for "beast" in Hebrew letters gives both numbers, and Irenaeus, the writer closest to the text, knew no Nero reading.

The English ciphers. No scholarship treats them; the criticism comes from inside the practice. Bethsheba Ashe, who makes a Hebrew calculator, argues that Agrippa never used his table, that calling it a Jewish method is false and feeds antisemitic conspiracy talk, and that the community's ciphers have no texts behind them. The practitioner Joachim Bartoll reports the community's own explanations of the names.

Isopsephy and Pythagoras. No early source has Pythagoras number letters, and his biographers wrote centuries after him; the claim that Greek buildings encode names rests on nothing we could find. Psychoyos's study of isopsephy, known to us through citations, places the practice in the Roman period.

Our own test. In October 2026 we checked the engine on 76 strings against a textbook implementation, open-source packages and three calculator sites, and the Hebrew methods against native calculators. Every table was right, but three things were wrong by design: Agrippa's table was missing, five mirror ciphers were shown as if they were equals, and transliterated values matched no tool. All three were changed, and our earlier page's claims of matching across four scripts were withdrawn.

Terms

Gematria: The value of a word as the sum of its letters, and the practice of reading words of equal value together. The Hebrew word comes from Greek, probably geometria or grammateia; scholars disagree.
Isopsephy: The Greek practice, from isos "equal" and psephos "pebble", the counter used for reckoning: words that come to equal pebbles.
Abjad: The old order of the Arabic letters, named after its first four (alif, ba', jim, dal), and the number values that go with it, 1 to 1000.
Mashriqi and Maghribi: The eastern and the western orders of the abjad. They give six letters different values; a word without them scores the same in both.
Numeral letters: Stigma or digamma (6), koppa (90) and sampi (900): Greek letters kept only as numbers after they had left the alphabet.
Final letters: The five Hebrew letters with a separate form at the end of a word (kaf, mem, nun, pe, tsadi). They count as their base letter, except in Mispar Gadol, where they are 500 to 900.
Digital root: A total reduced to one figure by adding its digits again and again: 376 gives 16, then 7.
Compound: In Cheiro's Chaldean table, the two-digit total read before its single digit, as in our Numerology chart.
Equivalence: Another word or name with the same value. The tradition reads such pairs together; it does not take them as proof of anything.
Asmakhta: Aramaic, "a support": the Talmud's word for a reading, such as a gematria, that supports a teaching without proving it.
Chronogram: A phrase whose letters add up to a date, the year of the event it records; common on Hebrew epitaphs and in Arabic and Persian verse.
Letter substitution: Exchanging each letter for another by a fixed rule before counting, as Atbash and Albam do. A substitution cipher of this kind is not itself a sum.
Core and more: Our chart's two tiers: the core ciphers are shown at once, the others behind "more ciphers".
The number of the beast: 666, or 616 in some manuscripts, in Revelation 13:18. Its usual scholarly reading is Nero Caesar in Hebrew letters, which remains a reading, not a proof.

Sample reading

Sample reading, AI-written for a fictional person.

Read the sample (1,189 words)

AI output can be wrong.

Jason, your name in full, Jason Carlos Eusebio, comes to 203 under English Ordinal and 68 under Full Reduction, and both fall to the same single digit: 5. That digit is the spine of your chart. It returns again under Standard (986) and under Offset 36 (833). Four separate counts of your three names, four times the same root.

You are counted in fives from the front

Look first at what that agreement is and what it is not. Two of those four are the plainest ways the modern sites count a Latin name: English Ordinal, a modern cipher that takes each letter's place in the alphabet, and Full Reduction, the 20th century table that folds the alphabet into nines. Standard is the old units, tens and hundreds pattern laid over the 26 letters, untraced in print. Offset 36 is a modern cipher that lifts each letter's place in the alphabet by the same fixed amount, and this chart calls it only by that plain name. Part of the echo is plain arithmetic. Your name runs to eighteen letters, and Offset 36 lifts each one by the same amount, so its total keeps step with your Ordinal and lands where the Ordinal lands. The tradition would call the repetition a correspondence. It is not a hidden signal, and it does not have to be one to repay attention.

What it gives you is an image of consistency under changed conditions. Count you simply, count you by the old place values, count you folded down, count you shifted up: the root does not budge. Your sense of yourself does not reshuffle when the room changes, and that is the one quality in this chart you can lean on hardest. People who meet you in different settings will compare notes and find the same person. You may not even notice this about yourself, because you only ever see yourself from the inside, in order, first letter to last.

Read from the back, you settle into four

Now turn the alphabet around. Under Reverse Ordinal, a modern mirror cipher from the 2010s, your name is 283. Under Reverse Reduction, its folded mirror, it is 112. Both fall to 4.

Here is the first real tension in your numbers. Read forward, you resolve to 5 across four ciphers. Read backward, you resolve to 4 across two. The two directions do not agree, and they cannot: the mirrored count of any name is fixed by its length, so for your eighteen letters the forward and backward totals always sum to the same figure, 486. The 203 and the 283 are one quantity split in two. Whatever you gain on one side you give up on the other.

The tradition reads its devices as a way of holding two words together and asking what one says about the other. Held that way, your pair describes a person whose forward motion and whose reflection are drawn from the same fixed store. When you act, you act from the 5. When you look back at what you did, you look from the 4, and the backward reading is the larger number. You know the moment. You send the message, the one you drafted carefully, and within a minute you are reading it again from the bottom up, weighing each line as it now looks from the other side. The second reading is longer than the first was. It often feels truer, and it often is not.

The oldest table leaves nothing over

Of all your values, only one comes from a table with a printed history: Latin (Agrippa), the Latin table Heinrich Cornelius Agrippa set out in Book II of his occult philosophy, printed complete at Cologne in 1533. It is a Latin table, not a Hebrew one, whatever the older calculator sites called it. Under it your name is 1386, and 1386 reduces to 9.

Agrippa gave this table together with a rule for finding a person's ruling star from what remains when the name's number is divided by nine. Whether he used it himself is not recorded. Your name leaves no remainder at all. 1386 divides by nine cleanly, and the count closes on itself at the last place rather than stopping somewhere along the way. Trigonal (1656), a modern cipher built from the ancient triangular numbers, closes the same way, also on 9.

So the one historically rooted reading in your chart sits apart from the fives and fours of the modern counts. Where the modern ciphers show you as consistent and divided, the 1533 table shows you as complete: a sum that does not spill over. There is a dignity in that, and a quiet risk. A count with no remainder has nothing left to carry forward. You are drawn to finish things properly, to close a chapter so thoroughly that nothing hangs loose, and you can confuse that completeness with rightness. Sometimes a remainder is where the next thing begins.

Where the count runs away from you

The pull that defines your chart is between the fives that agree and the totals that swell. Look at the small, folded ciphers side by side: Septenary 64, Chaldean 65, Full Reduction 68. Three tables from three different hands (a 2010s community cipher, Cheiro's table printed in 1926, the American numerology table) and all three put you in the same narrow band in the sixties. Then look at the multiplied ciphers. The cipher that multiplies each letter's place by six, a modern invention whose Sumerian label is a borrowing with no Sumerian source, gives 1218. Squares gives 3109. Primes gives 642. The same eighteen letters, inflated by rules invented to make bigger numbers.

The tradition's own correction applies here with force. With enough ciphers, every total matches something. Charles Swindoll showed in print in 1985, using the very multiplication behind your 1218, that a striking number can be found in almost any name. The large figures in your chart are where significance gets manufactured, and you have a temperament that feels the pull of manufactured significance. The 4 that reads you backward is the part of you most likely to go hunting through the big numbers for a meaning that confirms what you already fear or hope. That is the trap: mistaking the size of a number for its weight. Your steadiest values are your smallest ones.

Cheiro's table adds one structural note. Your Chaldean total, 65, reads under his compound method as 11, then 2: a doubled digit resting on a single one, two of the same standing together before they settle. It sits alone among your values, matching none of the others. Even in the one place your chart is paired, it reaches its root by a different road from everything else.

What to do with 203

The Talmudists called this kind of reading a support, not a proof, and that is the posture to take toward your own name. Take your Ordinal, 203, and its mirror, 283, and write them side by side on one page. Under the first, write one thing you did this week from conviction. Under the second, write what you told yourself about it afterward. Do this for a few weeks, and when the backward column starts outrunning the forward one, as your numbers suggest it will, let the 5 have the last word. Stop at the sixties. Leave the thousands alone.