Japanese

First place, alone.

--:--:-- [PST]
89.2
score for the human translation into Japanese — first, with no machine system alongside it
WMT25
72.5
Japanese score on real client content, among the lowest of eighteen languages
Engine index, 2026
Both agree
benchmark and production data point the same way, which is unusual
WMT25 / engine index
S01

What the evidence says

In the 2025 evaluation, the human translation into Japanese scored 89.2 and ranked first on its own — no machine system reached the top group with it. The best model scored below it.

On real commercial content the picture is worse for machines, not better: an index of over five thousand evaluations across ninety-seven client projects puts Japanese at 72.5, among the lowest scores of any language measured.

Both datasets agree, which is unusual and worth noticing.

S02

Why it happens

Japanese grammatically encodes the relationship between speaker and listener — who outranks whom, how formal the setting is, how direct it is acceptable to be. English does not carry that information anywhere in the sentence. So the machine has to infer something the source never stated, and it guesses.

A 2026 peer-reviewed study of business emails translated into Japanese found that a plain "translate this" instruction produced only limited adaptation toward Japanese conventions, failing specifically on honorifics, formal address, indirectness in hierarchical contexts, and the appropriate directness between a superior and a subordinate. The study's finding was that supplying context in the instruction fixed it — which is to say the fix is human expertise applied up front, not a better engine.

S03

What it means for your workflow

A Japanese file that is wrong in this way is not wrong in a way a scoring model detects. It is grammatically correct, fluent, and rude. The reader who notices is your customer.

S04

What we do

We review for register and honorific consistency against the audience and the relationship, not just against the source. Where you are producing Japanese at volume, we would rather fix the instruction than the output — so we report the pattern, not only the instance.

S05

Send us Japanese output and tell us who it is addressed to

The relationship between writer and reader is the thing being got wrong, and it is not in the source text. Tell us, and we will check against it.

Send us a file to evaluate