The authors consider the features of multi-agent modeling for traffic optimization in the central areas of cities. While evaluating the unique challenges associated with the high concentration of vehicles, pedestrians and historical buildings, the potential of multi-agent systems to effectively solve the problem of congestion, safety and quality of life in urban areas is investigated. The potential of multi-agent modeling in the context of traffic management in the central areas of the city allows us to identify the key challenges and opportunities. Many scientists address the main aspects of such modeling and use them in the transport and road sectors. A review of current research and development has shown that multi-agent models aim to simulate and optimize the supervision and control of transportation in various traffic scenarios. Modeling traffic organization in the central areas of cities is one of the main elements of urban development planning and management. Due to the growing population of cities and the increasing number of vehicles, the problems of congestion, air pollution, and inefficient use of infrastructure are becoming increasingly relevant. Therefore, it can be noted that multi-agent traffic modeling opens up new prospects for developing effective traffic management strategies, providing a flexible and adaptive solution to these problems. The research analyzes the existing approaches, identifies the system`s key components, and develops a model that demonstrates the interaction between agents and the environment based on a mathematical description. A practical simulation of the model, carried out using the AnyLogic software on the example of Lesia Ukrainka Boulevard in Kyiv, confirms the effectiveness of the multi-agent approach. The results of the study indicate the possibility of applying the developed model to improve intelligent information systems for traffic flow management, which opens up new prospects for improving traffic in the central areas of cities.