Finished: version 1.3, corrected October 2026
Antikythera Cosmos
The oldest geared computer we know of, rebuilt from its tooth counts and drawn to the 2021 reconstruction by Tony Freeth's team at UCL: sixty-nine bronze wheels and the whole Greek cosmos on its front, turned by one crank in your browser and held against the real sky.
What came out of the sea
In 1901 sponge divers working a Roman-era wreck off the island of Antikythera brought up a lump of corroded bronze. It lay unregarded until May 1902, when Spyridon Stais, visiting the museum his cousin Valerios directed, noticed a gearwheel in it. The ship had gone down around 70 to 60 BC.
What survives is 82 fragments, about a third of the machine, in the National Archaeological Museum in Athens. Thirty gears are still in them: 27 in Fragment A and one each in B, C and D.
Reading through the corrosion
In 2005 an X-ray scanner was shipped to Athens and the fragments were scanned slice by slice. Inside the crust were the gears, their teeth countable where the rims survive, and thousands of letters of Greek, many read for the first time since antiquity.
From those counts, and from the numbers inscribed on the front cover, Freeth and his colleagues worked out the rest. Thirty-nine of the 69 wheels here are their proposals, not readings, and the exhibit marks which is which.
The cosmos on the front
The front is the 2021 model's great proposal: the whole Greek cosmos on one dial, drawn here to the proportions of the paper's Fig. 7. The dome of the Earth sits at the centre. Around it run rings for Mercury, Venus, the true Sun, Mars, Jupiter and Saturn, each planet marked by a coloured stone, then a narrow date ring, the zodiac and the Egyptian calendar.
The Sun is the "little golden sphere" with its "ray" that the inscription on the back cover describes. Its ring carries the Age of the Moon, read by the Moon pointer, and small triangles for the eclipse limits: an eclipse is possible when the Dragon Hand, pointing to the Moon's nodes, lies inside them at new or full moon. With Aries at the top, each quarter of the zodiac sits beside its column of the star calendar on the plates above and below.
Index letters on the rings
The front cover carried an inscription listing, planet by planet, the events of each synodic cycle: conjunctions, greatest elongations, stations and oppositions. The 2021 team propose that each ring was marked where the Sun pointer stands at each event, with a letter beside the mark keying it to its line of the inscription, one alphabet running from Α on Mercury to Ω on Saturn.
The paper gives no angles for the marks, so, as it suggests, the machine is its own division engine: each mark here is cut where this copy puts the Sun on that ring at the event, found by turning it through forty years. Mercury's first three letters are cut faint, because those lines of the inscription are lost.
The Moon's phase
Fragment C preserves part of the Moon's phase device, with its small crown gear; the ball it turned, half silver and half black, went round once a lunar month. In the 2021 model the mean Sun comes up its own tube beside the lunar arbor to a 20-tooth wheel, b0, which turns that crown gear as the Moon carries it round; the gear's axle runs out along the Moon's arm to the ball.
Because it is turned by the difference between the Moon and the mean Sun, the ball shows the phase: black facing out at new moon, silver at full.
Exact in the teeth
Every gear is cut from one table of tooth counts, meshes and couplings, and every rate is an exact fraction, checked by tests against the 2021 paper. The Moon's pointer runs through 64 on 38, 48 on 24 and 127 on 32, which multiply out to exactly 254/19: in nineteen years the Moon goes round the zodiac 254 times.
| Planet | Synodic cycles | in years | Where the numbers come from |
|---|---|---|---|
| Mercury | 1,513 | 480 | derived by Freeth et al. |
| Venus | 289 | 462 | 462 is read on the front cover |
| Mars | 133 | 284 | derived |
| Jupiter | 315 | 344 | derived |
| Saturn | 427 | 442 | 442 is read on the front cover |
The crown wheel
The machine was driven from the side. A crown wheel, a1, of which the keyway for a removable crank survives, stands on the front face of the great four-spoked wheel b1 at its rim, its 48 teeth meeting b1's 223.
About four and two-thirds turns of the crank carry b1, and with it the mean Sun, once round: one year. Every other pointer on the machine is geared off that single rotation.
The pin and slot
The Moon does not move at an even pace: it hurries near perigee and lags near apogee. At the back of the machine, two 50-tooth wheels sit face to face on axes 1.1 mm apart, a pin on one riding in a slot on the other. As they turn, the slotted wheel speeds up and slows down by about six and a half degrees.
The pair rides on a platform that turns in just under nine years, so the swing follows the slowly turning line of the Moon's orbit.
Calendars on the back
The upper spiral is the Metonic calendar: 235 months in five turns, nineteen years. As the 2008 study reconstructs it, seven of the years have a thirteenth month, each year opens with the sign for "year" and its number, and numerals for the days skipped to keep the months in step are cut inside the first turn. The month names belong to a Corinthian calendar, most likely that of Epirus.
Inside it the Games dial counts the four-year round of the games, its years inside the rim and the games outside: the Isthmia, Olympia, Nemea and Pythia, with the Naa at Dodona and the Halieia of Rhodes. A second, conjectural dial opposite it, of which no trace survives, is given to the 76-year Callippic period. Each spiral pointer is a bar sliding through a guide on its arbor, so that a follower under its tip rides the groove and the tip always points to the current month.
Predicting eclipses
The lower spiral is the Saros, 223 months after which eclipses repeat. In Freeth's 2014 reconstruction 51 of its cells carry glyphs, 18 of them surviving: Σ for an eclipse of the Moon and Η for one of the Sun, 38 lunar and 28 solar, with the hour on a twelve-hour scale and an index letter, the letters running in alphabetical order through the glyphs to key them to the eclipse inscriptions.
A Saros runs about a third of a day past a whole number of days, so the small Exeligmos dial adds 0, 8 or 16 hours to the times. The exhibit holds each glyph against NASA's Five Millennium Canon of eclipses.
Checked against the sky
The exhibit keeps a second column beside the machine: where the real Sun, Moon and planets stood that day, from a modern ephemeris, and which eclipses really happened, from NASA's Five Millennium Canon. When the bronze is wrong, it says so.
- 28 of 28
- solar glyphs in one Saros from the epoch fall in a month with a solar eclipse somewhere on Earth
- 27 of 38
- lunar glyphs land on a real lunar eclipse; 10 more on one that only grazes the Earth's outer shadow, and one on none
- 3.5°
- the Moon stays within about this of the sky near the epoch, then drifts about a degree every eighteen years
- 54°
- how far Mars strays within ten years: the planets keep the inscription's period relations, but one epicycle each cannot follow their loops, and the machine spaces them evenly where the sky does not
Every comparison with the papers is written up: the first accuracy pass, the second, which fixed a Moon pointer running up to 20° off and a broken second epoch, and the third, which held the model against the papers' own figures and rebuilt the front display, the phase drive and the back dials to match them.
One crank, one year
A hand crank on the side turns a 48-tooth crown wheel against the 223-tooth main wheel. One turn of the main wheel is one year, and every other pointer is geared off it. Here the front dial runs through the year from 12 May 205 BC, the full moon of the eclipse dial's first month, and stops at each month whose Saros cell carries an eclipse glyph, so the glyph can be held against NASA's canon.
The gold ray from the golden sphere is the Sun, on its own ring; the silver hand is the Moon, with its half-dark phase ball riding on it; the stones on the rings are the five planets, and the dark hand that crosses the whole dial is the Dragon, pointing to the Moon's nodes. Drag the crank below to turn it yourself.
12 May 205 BC
0.00 turns of the main wheel
Saros cell 1 on the back
Where this copy departs from the 2021 paper
Every tooth count, gear train, pin-and-slot offset and pin distance follows Freeth et al. 2021 (Fig. 3 and Supplementary Tables S8 and S9), and the front display follows its Fig. 7 and the index letters of Fig. 8. The paper does not give the angles of the planets' scale marks, so, as it proposes, they are found with the machine itself: each is cut where this copy puts the Sun on that ring at the event. The back dials follow Freeth et al. 2008 and Freeth 2014, with the Games dial's Halieia from Iversen's later reading (2017). These parts are this copy's own choices:
- The size of the Sun's anomaly. The paper gives the true Sun's gearing (56, 52, 56 and a follower) but not its dimensions. This copy sets the pin for an eccentricity of 1/24, the ancient value that Ptolemy gives from Hipparchus.
- How the pointers are set. The paper says each anomaly device was set against observations on chosen dates. This copy sets each pin's starting phase, and each pointer's zero at the epoch, by fitting a modern ephemeris (astronomy-engine), which no ancient maker could have done. The 2021 paper does not choose an epoch; the two offered here come from Carman and Evans and Freeth (both 2014) and from Voulgaris and colleagues (2023).
- Where the calendars start. Nothing fixes which year of the Metonic cycle the maker began from. Here a cycle begins with the first new Moon after the autumnal equinox of 221 BC, the Corinthian new year, so that with the Games pointer set straight down at the start of the cycle, as the 2008 reconstruction proposes, the dial's Olympia fall in real Olympic years. The repeated month of the seven 13-month years is taken to be Eukleios, the 2008 study's "simple hypothesis", one of many possible. The Callippic dial, for which no evidence survives, counts its quarters from the cycle nearest Callippus's first period of 330 BC.
- Geometry, not ratios. The reconstructed wheels are cut at a module of 0.5 where the paper adjusts a few (about 0.53 and 0.54 for the 51 and 44 of the Venus train, 0.521 for its 34 and 26); arbor angles are this copy's own; and two surviving lunar wheels, e6 and k2, are cut at a module of 0.534 rather than the measured 0.536 and 0.541 so the train closes on its axis. The back dials are drawn smaller than the bronze to fit this copy's case. None of this changes a tooth count or a rate.
- A handle to turn it. The crown wheel a1 has a keyway for a removable crank; here a handle stays on its arbor so the machine can be wound by hand.
Two editions of one exhibit
Both carry every feature: a fourteen-step narrated walkthrough, the gear trains one at a time, the Fragment A scan, two proposed epochs, and a share link for any date you crank to.
How it was made
One gear table
Every tooth count, mesh, coupling and pin offset from the 2021 paper and its supplementary tables goes into a single file. Nothing else defines a gear.
Exact arithmetic
A Python solver turns the table into rates as exact fractions and lays out the arbors; tests hold each planet's period relation to the paper.
Built in Blender
Blender 5.1 cuts every wheel, plate and pointer from the table by script, with dial faces drawn to the millimetre, and exports the machine as glTF.
Driven in the browser
three.js re-derives every rate from rules carried in the model, so the browser turns the machine with the same arithmetic. Time runs as Julian Day on the proleptic Julian calendar, never a JavaScript date.
Checked, and corrected
Two accuracy passes held the numbers to the paper, one refereed by a second AI model whose every finding was verified against the sources. After corrections from Tony Freeth, a third held the model against the papers' figures, the 2008 and 2014 reconstructions included, and rebuilt what did not match: the front display, the crown wheel, the Moon's phase drive, the order of the gear layers and the back dials. The narrated walkthrough was filmed and critiqued, and its staging fixed where the picture missed what the voice described.
Designed and directed by Stewart Gregerson. Built with Claude Code. Narration in a synthetic ElevenLabs library voice.
Sources and credits
- Freeth, Higgon, Dacanalis, MacDonald, Georgakopoulou and Wojcik, A Model of the Cosmos in the ancient Greek Antikythera Mechanism, Scientific Reports 11:5821 (2021), CC BY 4.0. The reconstruction this exhibit follows, with the Antikythera Mechanism Research Project's papers of 2006 and 2008.
- Freeth, Eclipse Prediction on the Ancient Greek Astronomical Calculating Machine Known as the Antikythera Mechanism, PLOS ONE 9 (2014): the eclipse glyphs and the epoch. Carman and Evans, Archive for History of Exact Sciences 68 (2014): the epoch of 12 May 205 BC. Voulgaris, Mouratidis and Vossinakis, Almagest 14.1 (2023): 22 December 178 BC.
- Bitsakis and Jones, The Front Dial and Parapegma Inscriptions, Almagest 7.1 (2016), CC BY-NC 4.0. Iversen, Hesperia 86 (2017): the games and the calendar.
- Budiselic et al. 2020 and Woan and Bayley 2024: the calendar ring's holes. Edmunds, Journal for the History of Astronomy 42 (2011), and Szigety and Arenas 2025 (a preprint): the teeth and their errors.
- Eclipse Predictions by Fred Espenak and Jean Meeus (NASA's GSFC), the Five Millennium Canon. Positions from astronomy-engine by Don Cross, MIT.
| What | Who | Licence |
|---|---|---|
| Reconstruction model | Freeth et al., Sci. Rep. 11:5821 (2021) | CC BY 4.0 |
| Fragment A CT scan | Ashkan Pakzad | CC BY 4.0 |
| Parapegma and front dial inscriptions | Bitsakis and Jones (2016) | CC BY-NC 4.0 |
| Moon face | NASA CGI Moon Kit (LROC) | Public domain |
| Ephemeris (astronomy-engine) | Don Cross | MIT |
| Rooms (studio_small_09, artist_workshop) | Poly Haven | CC0 |
| Layout reference | Thomas Weibel | CC BY |
| Narration voice, David (Documentary) | ElevenLabs library | ElevenLabs terms |
The machine only knows turns. Give it some.