This project converts speech into text. At the run time speech data is acquire from microphone and converted into text speech frames the speech frames are then pass for preprocessing and after preprocessing of the sample frames HMM-based training is applied on speech frame. Functionally the project divided into three modules