<![CDATA[Speaking competence is a central pivotal aspect of English as a Foreign Language (EFL) instruction, particularly in higher learning institutions where students are expected to achieve communicative proficiency for academic and professional purposes. Despite its importance, EFL speaking instruction often faces challenges such as limited student engagement, insufficient opportunities for authentic interaction, and fragmented integration of emerging technologies. Previous studies have examined gamification, Artificial Intelligence (AI), and computational thinking separately, yet few attempts have been made to integrate these elements within a unified pedagogical framework. Therefore, this study aims to develop and validate a theoretical model for AI-integrated gamified EFL speaking instruction. The study employed a Design-Based Research (DBR) approach encompassing six operational phases. Data were gathered through literature synthesis, expert validation checklists, construct validation questionnaires, and pilot classroom observations. The expert validation results demonstrated high conceptual clarity and pedagogical alignment, yielding a robust overall mean score of 4.60. Furthermore, Confirmatory Factor Analysis (CFA) confirmed acceptable construct validity and internal consistency, with strong model fit indices (Chi-square/df = 1.87, CFI = 0.93, TLI = 0.91, RMSEA = 0.067). Pilot implementation further verified the practical feasibility of the model with an overall mean score of 4.36. This study offers a validated pedagogical framework that strategically synthesizes gamification, AI-supported feedback, and computational thinking to optimize EFL speaking environments in higher education.]]>
