Skip to content
SourceTarget
Quality metrics

BLEU score

BLEU (bilingual evaluation understudy) is an automatic metric for machine translation quality. It scores a translation from 0 to 100 by measuring how much its wording overlaps with one or more human reference translations of the same text. Higher overlap means a higher score.

How it works

BLEU compares the machine's output against a human reference and counts matching runs of words. It rewards output that reuses the reference's exact phrasing and penalises output that strays, then folds those matches into a single number.

Because it only looks at word overlap, BLEU is fast, cheap and perfectly repeatable: the same input always produces the same score. That is also its limit. A translation can say the right thing in different words and score poorly, or match many words while mangling the meaning and score respectably.

How SourceTarget uses it

SourceTarget calculates BLEU as one input among several, never as the final verdict. It is useful for comparing engines on the same content and for tracking whether a change helped or hurt.

On its own it cannot say whether a single sentence is fit to publish, so it feeds the composite quality score alongside metrics that judge meaning, not just words, and the result routes work to the right level of human review.

BLEU score compared with COMET score

BLEU scoreCOMET score
What it measuresWord overlap with a referencePredicted human judgement of meaning
StrengthCheap, transparent, repeatableCorrelates far better with human ratings
WeaknessBlind to correct paraphrasesA model prediction, weaker in low-resource languages
Best used forComparing engines quicklyJudging adequacy and fluency

Common questions

99%
certified-job accuracy, human-reviewed at every quality gate.
70%
faster turnaround, pairing an AI production line with NAATI linguists.
300+
languages quoted and delivered, including First Nations languages.
Language Capability Statement
PDF · ISO 9001 · 17100 · 27001 · NAATI endorsed
DOWNLOAD
Census Language Data Pack
PDF · ABS-grounded, suburb-level coverage
DOWNLOAD
Paste your text. Get a price in seconds.
Certified documents, web content, campaigns: quote any job instantly, in 300+ languages.
GET AN INSTANT QUOTE
SourceTarget
The language layer for multilingual Australia. Census data for every suburb, NAATI certified translation in 300+ languages.
ISO 9001ISO 17100ISO 27001NAATI ENDORSED
ACKNOWLEDGEMENT OF COUNTRY
SourceTarget acknowledges the Traditional Custodians of the lands on which we live and work, and pays respect to Elders past, present and emerging. We recognise the enduring connection of Aboriginal and Torres Strait Islander peoples to language, culture and Country, and the hundreds of First Nations languages that remain a living part of Australia today. Sovereignty was never ceded.
SourceTarget delivers NAATI certified translation and localisation services for Australian government, health and not-for-profit communicators. From certified document translation and website localisation to transcreation and emergency multilingual communications, every project is grounded in ABS Census language data and reviewed by NAATI certified translators. Explore the Australian Language Map to plan campaigns suburb by suburb, across 300+ community languages in 551 LGAs and 4,288 suburbs.
© 2026 SourceTarget Pty Ltd · ABN 00 000 000 000 · Australian translation services and multilingual communication.