<p>Over the past few years, social robots have emerged as potential tools for enhancing human well-being, leveraging their advanced cognitive abilities to provide emotional support and resource distribution. The main objective of this research is to empirically investigate how the cognitive capabilities of social robots affect human well-being. Specifically, this study examines whether anxiety and mental well-being correlate with a preference for and satisfaction from interactions with emotionally intelligent and resource-efficient social robots. Data collection combined quantitative and qualitative approaches. There were 1,000 structured questionnaires distributed online and offline to evaluate emotional states pre- and post-interaction with social robots in various regions of China, using psychological scales and a 5-point Likert scale. Additionally, semi-structured interviews gathered qualitative insights. With a 90.01% response rate from 865 valid responses, preliminary analysis showed a significant correlation between increased anxiety levels and social robots’ need for emotional support. Participants with greater mental well-being preferred emotionally intelligent robots, with notable satisfaction and engagement among users interacting with robots having advanced cognitive abilities. This research makes a novel contribution to the field by empirically demonstrating how the cognitive skills of social robots can influence human emotional and psychological well-being. The study provides a comprehensive understanding of human interactions with social robots, by assimilating quantitative data with qualitative insights. It highlights the essential role of advanced cognitive functions in social robots, significantly enhancing their ability to support human well-being and pushing the boundaries of human–robot interaction research.</p>