Documentation Center

Support for Machine Translation Quality Estimation (MTQE)

With the evolution of Generative AI and Large Language Models (LLMs), machine translation providers such as RWS Language Weaver or ModelFront have introduced Machine Translation Quality Estimation (MTQE). This new feature assesses the quality and reliability of machine translation output, enabling you to determine when human editing is needed.

To support this enhancement, Trados Studio has extended its machine translation capabilities, enabling you to check MTQE results directly in the Editor view to help streamline your post-editing efforts. The MTQE results are preprocessed by machine translation providers and displayed in Trados Studio in the new TQE panel.

Translations are color-coded as green, amber, or red based on accuracy, allowing you to focus on segments that need attention. You can also filter the results by color code to prioritize your workflow.

MTQE is available for cloud projects, but you can also use it offline by downloading preprocessed SDLXLIFF files. The only limitation to this method is that you will no longer have color coding on your results. To use MTQE at the time of writing, you need a subscription to a machine translation provider with quality estimation, custom workflows based on Trados Enterprise or Accelerate, and Trados Studio as a desktop client and online editor.