Convert Roman numerals — both directions at once
Type a number from 1 to 3999 on the left or a Roman numeral on the right — the counterpart appears instantly, along with the place-value breakdown. The clock-face form IIII is read and flagged; output is always the standard spelling. Invalid input such as IIX is rejected with a stated reason instead of a guess.
How does the Roman numeral system work?
It gets by with seven symbols: I (1), V (5), X (10), L (50), C (100), D (500) and M (1000). Each place value is written on its own and the parts are joined. 2026 consists of two thousands (MM), two tens (XX) and six ones (VI), so MMXXVI. A smaller symbol in front of a larger one is subtracted — but only in six allowed pairs: IV (4), IX (9), XL (40), XC (90), CD (400) and CM (900). 1994 therefore reads as M + CM + XC + IV. That is exactly the breakdown the tool shows below the result, place by place.
Why is IIII common on clock faces — yet not standard?
The Romans themselves often wrote four additively as IIII; the strict subtractive rule taught today only became a binding convention long after antiquity. On classic clock dials the additive form survives to this day — the Great Clock at the Palace of Westminster, with its IV, is the famous exception. This tool draws a clear line from that: IIII, VIIII or XXXX are accepted when reading and flagged as non-standard, with the standard form given as the answer. They are never written out — otherwise the converter would carry the deviation forward.
Which inputs does the tool reject — and why?
Anything that does not resolve to one unambiguous number is turned down with a named reason: VV is forbidden because V, L and D never repeat (two of them would be the next larger symbol). IL is not an allowed subtraction — only I, X and C may be subtracted, and only from the next two larger symbols, which is why 49 is XLIX. IIX fails on ordering, because 8 is VIII. A wrong number would be worse than an honest error message, so the tool never guesses.
Why does the range end at 3999?
Because the classic notation has no symbol above M (1000) and allows at most three identical symbols in a row — the largest representable number is MMMCMXCIX. For anything bigger, scribes used the vinculum, an overline that multiplies a numeral by 1000. That is a notation of its own which plain text cannot carry reliably, so at 4000 the tool honestly says this is where it stops.
What about Roman numerals in Unicode?
The Unicode standard has dedicated code points for Roman numerals in its “Number Forms” block: U+2160 through U+2188, including single characters such as Ⅻ for 12 and ↈ for 100,000 (source: Unicode Consortium, “Number Forms” block). Hardly anyone types those in normal text, though — Roman numerals are written with the letters I, V, X, L, C, D and M. Those are exactly what this tool reads, in upper or lower case; a pasted numeral character such as Ⅳ is called out as a foreign character instead of being silently swallowed.
Does my input stay private?
Yes. The conversion is a handful of lookups and additions that run entirely in your browser — no request leaves the page, and it also works offline. There simply is no server that could log what you type. You can still share a result: the input is kept as a parameter in the address bar, so a copied link shows the same conversion.