<p>Live streaming platforms have increasingly reshaped how audiences interact with content creators, blending entertainment with real-time engagement. This study categorizes audience behaviors that demonstrate support for streamers into three types: non-creative, contributory, and monetary participation. It simultaneously considers the effects of streamer characteristics, content information quality, and platform user experience on these behaviors. Adopting a para-social interaction perspective, the study examines how perceived expertise, trustworthiness, visual attractiveness, and content quality influence the development of para-social interaction, and how such interaction, in turn, drives the three types of audience participation intentions. Furthermore, the study explores the moderating role of cognitive load in the relationship between para-social interaction and participation behavior. Data were collected through an online survey (N = 718) and analyzed using SPSS and AMOS. The results reveal that para-social interaction exerts a positive influence on all three forms of participation intention. Additionally, expertise, visual attractiveness, and content quality significantly enhance para-social interaction, while trustworthiness does not show a significant effect. Cognitive load was found to negatively moderate the relationships between para-social interaction and non-creative as well as monetary participation intentions. These findings contribute to a deeper understanding of audience engagement in live-streaming contexts and offer practical insights for improving streamer strategies and platform design.</p>