Nevisa dictation system is a large-vocabulary speech dictation system based on ASR Gooyesh continuous speech recognition engine. Two languages, Persian and English are supported and recognized instantly without dependence on speaker. Thus you can save time and cost significantly in writing documents either in Persian or English. The flexibility of Nevisa in adverse environmental conditions is achieved by advanced noise robustness techniques. The background adaptation makes the recognition more accurate as the time goes on. By now the accuracy of Nevisa dictation system is more than 85%.