Cost control in construction projects is a core aspect of engineering management, yet achieving precise control is challenging due to various uncertainties. To address this issue, this paper proposes a genetic algorithm-based model for optimizing construction project costs. The model quantifies building data features as algorithm inputs and utilizes evolutionary strategies such as selection, crossover, and mutation to iteratively optimize design schemes and resource allocation, ultimately minimizing total costs. Based on this algorithmic principle, we have developed a comprehensive software system applied in real projects. Test results demonstrate over 10% optimization across multiple cases, significantly reducing construction costs. The system exhibits high computational efficiency and interpretability, supporting efficient decision-making and result analysis. This study provides a robust technical means for lean cost management in the construction industry.