The current state of freight transport infrastructure in Indonesia lacks a comprehensive multimodal perspective. Developing infrastructure within this perspective is crucial to achieving efficient and seamless freight transportation. A strategic model is proposed to optimize benefits by minimizing total distribution costs while adhering to budget constraints for investment, operation, and maintenance (IOM). The model incorporates rail, road, and sea modes, factoring in generalized transport costs, such as the value of time (VOT) and reliability (VOR), under various scenarios. Applied to Java's freight transport network, this model uses network representation to reflect real conditions, including attributes like speed, travel time, unit costs, and capacity. It predicts freight volumes for a target year and employs bilevel programming solved by a genetic algorithm to determine the best infrastructure development strategies. Results demonstrate that a multimodal approach yields lower total distribution and IOM costs and higher efficiency compared to a unimodal approach. The model serves as a tool for optimizing multimodal transportation systems, reducing generalized costs, and providing strategic recommendations for infrastructure development in Java. © 2025, Horizon Research Publishing. All rights reserved.