Objective: To end AIDS epidemic by 2030 despite the new challenges brought by Covid-19, such as increasing poverty rates and inequalities, policies should be designed to deal with population heterogeneity and environmental changes. Bottom-up designs, such Agent-Based Model (ABM), have the potential to deal with such complexity by explicitly model these features. HIV/AIDS have a complex dynamic by structural factors, risk behaviors, biomedical characteristics, and interventions. All this embedded into a social structure with inequalities, stigma, and prejudice. To understand how this complexity is faced with ABMs, we performed a scoping review of its applications for HIV to highlight the gaps and potentialities of such approach. Methods: We search PubMed, Web of Science and Scopus repositories, following the PRISMA extension for scoping reviews. We identified the main articles by a co-citation network, and categorize the literature Aims, (sub)populations, and countries/regions under study. Results: We found 73 articles that applied ABM to HIV. From them, we identified 16 main articles. Most of the studies model Transmission Dynamics (23/78 – relate to aim categories) into specific key populations (19/73 – related to papers) of US and South Africa (50/66 – related to country appearances). More recent studies applied ABM to model PrEP interventions (10/78 – related to aim categories) and Racial Disparities (9/101 – related to populations appearances). Conclusions: Availability of data, computational power and specific know-how are the major barriers to its diffusion. We also identified the literature is highly concentrate on researchers located on the United States with low levels of outsiders collaboration. This reveals how infant is the applications yet. Overall, ABMs are still underused considering their potentialities, and should be applied more broadly to support the implementation of HIV/AIDS control interventions aimed to end AIDS epidemic by 2030.