Nowadays, the task of preventing suicide is one of the priorities in the health sector. Therefore, it is important to identify people prone to suicide at an early stage. This article discusses the possibility of real-time detection of visited websites containing suicidal statements. The classification of web pages is based on the analysis of the text contained on it. This work can be divided into two parts: creating a browser extension and the server. The extension collects information about the content of the web pages visited by the user and transmits it to the server. The page classification process takes place on the server. In the final part of this work, a comparison of the effectiveness of detecting suicidal websites using various machine learning algorithms is presented.