The emergence of Large Language Models (LLMs) like GPT-4 revolutionizes mental health care, enhancing counseling and therapy. This study investigates the factors influencing user acceptance of Large Language Model (LLM)-powered conversational agents in mental health support, utilizing frameworks that integrating Technology Acceptance Model (TAM) and the Unified Theory of Acceptance and Use of Technology (UTAUT). It examines how perceived benefits, perceived risks, trust, and perceived anthropomorphism shape acceptance, with attention to gender differences and familiarity with LLMs. A questionnaire-based study of 202 participants assessed these factors, and multiple linear regression analysis and structural model identified performance expectations and social influence as significant predictors of acceptance, supporting TAM’s core concepts. The study concludes that perceived benefits are crucial for acceptance, with trust also playing a key role. Enhancing performance expectations can mitigate technological concerns. These insights are vital for designing effective LLM-powered mental health support systems, aiming to improve user satisfaction and overall mental health outcomes.