Machine Translation (MT) has made significant advancements due to the rapid expansion of natural language processing. Even though several machine translation methods have been released in recent years, there hasn't been enough focus on automated and intelligent quality detection for translation outcomes. Neural Machine Translation (NMT), a machine translation method that is data-driven, is more effective in large corpora but with restricted corpus resources, there is still a sizable range of opportunity for advancement. To overcome these issues, design a system module of sentence translator using Neural Machine Translation is presented. This analysis will construct a translation system depending on the GRNN (Gated Recurrent Neural Network) deep learning algorithm and comprises the creation of the attention, pre-processing, coding, and decoding modules. This system does the Language translation with improved Accuracy and performance.