This study aims to develop a conceptual framework for a numeracy assessment model grounded in local cultural contexts and supported by Computer Assisted Test (CAT). The proposed model addresses the persistently low numeracy proficiency among Indonesian students, as evidenced by national and international assessments such as PISA and the National Assessment, as well as the scarcity of culturally relevant and contextually meaningful assessment instruments. By integrating constructivist learning theory, ethnomathematics, authentic assessment principles, and item response theory (IRT), this framework is designed to render numeracy assessments more contextual, adaptive, and psychometrically reliable. The research yielded a comprehensive development model comprising the following stages: (1) planning through systematic literature review and field observation; (2) design of numeracy competency grids and culturally contextualized test items rooted in Bugis-Makassar traditions; (3) expert validation to ensure content and cultural validity; (4) digitization of instruments into a CAT platform; and (5) pilot testing to collect empirical data on item characteristics, reliability, and validity. This study is expected to make a significant contribution to enhancing students’ numeracy literacy by promoting authentic, culturally embedded assessments. Furthermore, it supports the broader adoption of digital assessment technologies in educational settings. By engaging students with assessment experiences anchored in their local cultural heritage, the model aims to foster numeracy competencies that are not only academically robust but also critically reflective, contextually grounded, and applicable to real-world challenges.