The aim of this study is to examine the role of the social network X in shaping public opinion regarding the 12-day war between Iran and the Zionist regime. The research is designed around four hypotheses concerning the existence of significant relationships among the aforementioned variables. It employs a descriptive-analytical method based on a quantitative content analysis of 1,000 selected posts from six user groups—government officials (231 posts), media (103 posts), analysts (129 posts), artists (29 posts), athletes (52 posts), and ordinary users (456 posts)—posted between June 13 and June 24, 2025. The study's primary variables include user group type, news framing (security-military, humanitarian, political-diplomatic, general), engagement levels (likes, reposts, comments), content orientation, and lexical highlighting techniques. Descriptive findings indicate that the security-military frame appears most frequently (34.8%), whereas political-diplomatic frames garner the highest average number of likes, reposts, and comments. Accounts belonging to government officials recorded the highest average engagement within the sample. Inferential findings reveal significant relationships between frame type and engagement level, account type and engagement volume, and content orientation and engagement level. Furthermore, an ANOVA test confirms a significant difference in average engagement across the various frames. In other words, user reactions in the studied sample were not solely determined by the volume of published messages; rather, the news framing, account type, content orientation, and certain lexical salience factors were associated with engagement levels. Lexical analysis revealed no significant relationship between the number of hashtags and user engagement, whereas the inclusion of key concepts in the text showed a positive and significant correlation with engagement. These findings indicate statistical associations among content characteristics, account type, and engagement within the research sample; however, they do not allow for the inference of causal relationships or the internal workings of the X algorithm.