This study explores AR technology's role in assessing EFL students' speaking skills, aiming to enhance language learning experiences. Its main objective is to design and validate an AR-based assessment module for evaluating speaking proficiency. Employing a mixed-methods approach, the research investigates AR's impact on student engagement, learning outcomes, and language proficiency. It involves forty-five EFL learners from IAIN Fattahul Muluk Papua, using both quantitative and qualitative methods to collect data, including pre-tests, post-tests, interviews, and observations. The AR-based module integrates technology into authentic assessment scenarios, offering immersive learning experiences. Results show a significant improvement in post-test scores, with the mean score increasing from 2.4 (pre-test) to 3.9 (post-test), indicating the module's efficacy. The paired-samples t-test revealed a significant difference (t (59) = -10.24, p .001) between pre-test and post-test scores. Qualitative insights highlight increased engagement, positive experiences, and enhanced confidence among students. These findings underscore AR's potential in language education, particularly in speaking assessment. The integration of AR-based modules presents a promising approach for enhancing language teaching and learning, providing immersive experiences. This has implications for educators and practitioners seeking innovative language assessment methods