Depression is a mood disorder which causes persistent feeling of sadness and loss of interest. According to a survey of World Health Organization (WHO), approximately 3.8% of the world population is affected (which is about 280 million people) by depression. Depression affects millions of people worldwide knowingly or unknowingly. Depression and mental illness are a key problem nowadays in the society. Many people will not be interested in expressing their feeling to others. In this case, texts can be a good option to analyse one’s emotion. The automated system that can help in detecting depression in people of various age groups will be a great advantage in this aspect to detect the depression. In this application, Artificial Intelligence (AI) and diverse Deep Learning (DL) techniques can be used to detect depression. Depression can be predicted in different ways by analysing the videos, speech or texts. Nowadays since everyone use to chat and use text-based messages and posts, those texts can be used to predict their emotion and predict depression. Before consulting a mental health expert, automatic depression detection through text messages can assist people confidentially and conveniently understand their mental health state. This work is to implement the depression detection using the multiple layers in RNN algorithm. Here using RNN algorithm, an accuracy of about 99.66% is obtained.