Project tools

CAT settings reference

Translation defaults, translation memory, terminology, quality checks, and presets.

Settings in Language assets hold matching, QA and job defaults: language and tone defaults for translation, and a separate domain default for terminology extraction.

What you can configure

SectionSettings
Translation defaultsDefault language pair and tone for new translations. A saved domain applies only to terminology extraction.
TM behaviorEnable TM, set the suggestion threshold, fill empty editor units with exact matches, save approved Translation Editor units, and propagate approved repeats to unedited units. Use TM must also be enabled for the document job.
Glossary behaviorCase sensitivity and word-form matching. Matched terms are supplied to the model; glossary QA flags missing target terms.
TerminologyAutomatic extraction after translation and whether candidates require review. Auto-approval enters the terminology library; it does not automatically add entries to a glossary.
QualityGlossary adherence, consistency, numbers and dates, inline tags, placeholders, untranslated text, punctuation, length, and forbidden terms. Each check can be enabled separately.
PresetsNamed bundles of settings you switch between, one per client or content type.

Presets in practice

A preset saves the language pair, translation tone, and glossary choice, plus separate terminology settings such as domain. For example, *Acme Legal* can use a formal translation tone and the legal domain for terminology extraction. Presets are applied in the upload dialog and saved from its summary (Save as preset).

Why the settings list is limited

The settings shown correspond to supported workflow behaviour. QA checks flag issues for review; they do not guarantee translation accuracy. Automatic locale conversion, brand detection, and a document-wide consistency lock are not offered.