Interpreting Mandarin Measure Words and Numbers Correctly, Automatically
Two specific patterns cause most of the numeric and grammatical errors in Chinese-to-English translation: measure words, which don't translate to anything and must be dropped, and large numbers, which are grouped differently in the two languages and need conversion, not word-for-word translation.
Measure words: nothing to translate, easy to mistranslate
Mandarin requires a measure word between a number and the noun it counts — 三个人 (three [measure word] people), 一位客户 (one [measure word] client), 两家公司 (two [measure word] companies). English has no equivalent requirement, so the correct translation simply omits it: "three people," not "three units of people."
The failure mode is translating the measure word as if it carried independent meaning, which produces stilted or outright wrong English. A system built for this language pair needs to recognize measure words as grammatical scaffolding to discard, not vocabulary to render.
Numbers: grouped differently, not just written differently
English groups large numbers in thousands: 1,000 / 1,000,000 / 1,000,000,000. Mandarin groups in ten-thousands: 万 (10,000), 亿 (100,000,000). A number like 850,000 is naturally spoken as 八十五万 — eighty-five ten-thousands — not as a direct digit-by-digit reading. Translating the spoken structure literally, rather than converting the underlying value, risks a garbled or order-of-magnitude-wrong figure.
This is a narrow, mechanical problem with a mechanical fix: convert the underlying numeric value explicitly, rather than trusting a general translation model to reliably regroup digits correctly in real time, under the added pressure of doing it live.
Why this deserves dedicated handling
Neither of these is a matter of translation "quality" in the usual sense — a model can be excellent at conveying meaning and still make both of these errors, because they're structural, not semantic. That's why a system built specifically for Chinese-English simultaneous interpretation benefits from handling measure words and number grouping explicitly, rather than leaving them to general-purpose translation.
Be first to try real-time interpretation yourself.
In active testing — join the waitlist and we'll email you when it's ready.
Read next
Chinese–English Interpretation: Why Simultaneous Translation Is Uniquely Hard
Chinese and English put information in different places in a sentence — Chinese often saves the key verb or qualifier for later than English would — which means a simultaneous system sometimes has to wait for more of a Chinese sentence before it can commit to an English translation, or vice versa.
What Is Real-Time Speech Translation, and How Is It Different From an Interpreter App?
Real-time speech translation listens to speech as it's spoken and produces translated text (and often audio) within a few seconds — continuously, not after you finish a sentence and hit a button.
Simultaneous Interpretation for Video Calls: How It Actually Works
On a video call, simultaneous interpretation is a short pipeline running continuously in the background: audio in, transcribed speech, translated text, and captions or synthesized speech out — fast enough that it never asks the conversation to pause and wait for it.