Learn about Generative Translation

Generative Translation harnesses a large language model (LLM) for translation by providing it with the necessary context and linguistic data to produce the best possible translation.

 

During project processing, Generative Translation consolidates linguistic resources – such as translation memories (TM), terminology and any machine translation (MT) provider - with the power of a LLM to deliver higher quality translation results. Armed with your linguistic data, the LLM automatically applies any relevant terminology from your termbases and aligns its output with previously translated segments available in your TMs, ensuring consistency with your established tone of voice.

 

To refine the translation further, you can add custom instructions – such as removing gender bias, specifying the target audience, or adjusting the tone of voice. These instructions help the LLM make decisions about the translation it generates. With this feature, project managers can reduce the need for extensive post-editing, speeding up time to market, improving quality, and increasing client satisfaction.