Road transport is a major contributor to urban energy consumption and greenhouse gas emissions, yet traditional navigation systems often prioritize travel time or distance, neglecting critical factors such as topography. In cities with hilly terrain, such as Divinópolis - MG, road gradients strongly influence fuel usage, meaning the shortest routes are not always the most efficient in terms of energy use. This work aims to develop and validate a deterministic eco-routing framework that minimizes estimated fuel consumption while maintaining competitive travel times, without relying on massive historical Floating Car Data (FCD). The proposed methodology utilizes OpenStreetMap data and digital elevation models to construct a routable graph enriched with a cost function grounded in vehicle dynamics that penalizes steep gradients. Four algorithms were implemented and compared: Eco-Dijkstra, Eco-A*, Fastest-Dijkstra and Fastest-A*, employing an admissible heuristic based on the Haversine distance to ensure optimality. Experimental results demonstrate that the ecological route reduces fuel consumption by approximately 5.44% compared to the shortest path. Notably, while the average travel time increased marginally by 1.79%, the analysis indicates that eco-routes can, in certain cases, maintain or even decrease travel duration. Furthermore, the A* algorithm outperformed Dijkstra in computational efficiency, reducing execution time by about 6.0%. This study concludes that integrating static topographic data into routing algorithms is a viable and scalable strategy for promoting sustainable mobility in medium-sized and developing cities.