The paper discusses digital technologies for building optimal routes in transport logistics. The goal of the work is to develop an effective algorithm for constructing optimal transport routes based on mathematical models on graphs. A model is built on a graph with nodes at the corresponding intersections of the road map. Based on the “A star” algorithm for finding the route with the lowest cost, a modified method is presented taking into account traffic jams and traffic rules at intersections in the weighting coefficients. Evaluations of the algorithm showed that the computational complexity of the presented algorithm is O((v + w)log v), where v,w is the number of vertices and edges of the model in the graph. In comparison with the classic Dijkstra’s algorithm with computational complexity O(v^2 + w), the presented algorithm is more optimal. Unlike Dijkstra’s algorithm, the “A star” algorithm works correctly for graph models with negative weighting coefficients. For a digital freight logistics service, a library with the presented algorithm has been developed in C# language. Using the developed library, an example of finding the optimal route on a road map using the presented algorithm is calculated. The developed algorithm and library can be widely used for digital freight logistics services.