From Pattern Recognizers to Personalized Companions: A Survey of Large Language Models in Mental Health

Abstract—The rising global prevalence of mental health conditions, together with longstanding barriers in traditional healthcare, such as limited resources, high cost, stigma, and privacy concerns, has created an urgent need for accessible and scalable support. Large Language Models (LLMs) have emerged as a transformative technology with strong potential to democratize mental health support through advanced natural language understanding and generation. However, the rapidly expanding, fragmented body of work in this area lacks a coherent evolutionary narrative, making it difficult to contextualize current progress and identify future directions. This survey addresses this gap by organizing and analyzing the literature around a central thesis: the role of LLMs in mental health is evolving through three distinct, increasingly sophisticated phases. We trace this trajectory from Phase I, in which LLMs act primarily as passive Information Tools and Pattern Recognizers for assessment; through Phase II, where they function as Empathetic Conversationalists for in-the-moment, stateless interactions; to the current frontier, Phase III, which seeks Longitudinal, Personalized Companions implemented as stateful cognitive agents.

Introduction. I. INTRODUCTION M ENTAL health has become an increasingly urgent global concern, with the prevalence of conditions such as depression, anxiety, and loneliness steadily rising across diverse populations [1], [2], [3], [4]. Traditional mental healthcare, while effective, is constrained by significant barriers, including resource scarcity, high costs, social stigma, and privacy concerns [5], [2], [4]. These limitations leave a substantial portion of individuals without timely access to support, often delaying intervention until symptoms become severe and outcomes deteriorate. The advent of Large Language Models (LLMs) [6], [7], [8] represents a paradigm shift, offering transformative potential to democratize mental health support. With their profound capabilities in natural language understanding and generation, LLMs have driven a new generation of mental health technologies. As these models evolve into multimodal LLMs (MLLMs), they can integrate non-verbal cues such as speech prosody and facial expressions, enabling more nuanced assessments and richer human-AI interactions.

Discussion / Conclusion. VII. CONCLUSION AND FUTURE DIRECTIONS This survey has mapped the rapid evolution of Large Language Models in mental health through a structured developmental lens. We traced the trajectory from Phase I, where models served as passive Information Tools for risk detection, to Phase II, where they evolved into Empathetic Conversationalists capable of conducting supportive, singlesession dialogues. Currently, the field stands at the frontier of Phase III, striving to engineer Longitudinal, Personalized Companions, stateful agents endowed with memory, planning capabilities, and tool use. While the progress is remarkable, the transition to fully autonomous, clinically valid agents remains incomplete. To bridge the gap between technological potential and clinical reality, future research should focus on the following critical directions: