The primary goal of our research is to automatically gauge how severe a user’s depression is based on their comments shared on social media, especially on Reddit. This detection process is designed to evaluate each user’s level of depression severity, which is divided into four categories: minimal, mild, moderate, and severe. Instead of just analyzing symptoms one by one, our approach takes a more comprehensive view by directly assessing the overall level of psychological distress reflected in users’ posts. To do this, we utilize cutting-edge natural language processing and deep learning techniques that can pick up on the linguistic signals tied to different degrees of emotional distress. By tapping into genuine and spontaneous text from users’ daily online interactions, our system allows for an automated, ongoing, and non-intrusive way to assess depressive states. Ultimately, this method could lay the groundwork for early detection and support tools for mental health professionals.