Aiming at the demand of Reactive Power Optimization (RPO) in this real power network, this paper uses the Java programming language to design and develop a RPO analysis software based on the Artificial Intelligence (AI) algorithms. The software integrates five AI algorithms with different properties, which can be used for RPO solution under different occasions. The user can quickly select the appropriate algorithm for RPO calculation based on actual situation, and the calculation results can guide the power grid company to optimize the parameters such as reactive power compensation capacity, transformer ratio, etc. The software mainly includes the project management module, data editing module, RPO calculation module and control variable editing module, and has BPA-Matpower data interface, which is compatible with BPA data format and can complete accurate conversion from BPA data to Matpower data. The software can directly edit the BPA data with different grid architectures without reprogramming, and has the characteristics of high precision and versatility. The effectiveness and practicability of the software is finally validated by an engineering case study, according to the BPA data of certain city grid under the China Southern Power Grid. This work is supported by National Natural Science Foundation of China (No. 51477055 and No. 51777078).