Micro-level features like property size, location, and amenities are essential for accurate prediction of real estate rental prices. These factors are often over-looked in macroeconomic-focused studies. In this study we implemented machine learning algorithms to predict properties rental price in Kuala Lumpur and Selangor using a dataset of over 19,000 properties. We have filtered the dataset, converts texts to numerical data, and normalized the dataset, before designing and training three artificial intelligent models: Support Vector Regression (SVR), Long Short-Term Memory (LSTM), and LSTM with Principal Component Analysis (PCA).). LSTM outperformed the others due to its ability to capture complex, non-linear patterns. LSTM-PCA, though slightly less accurate, offered computational efficiency. SVR performed the worst. The findings underscore the potential of AI in micro-level rental prediction and highlight the importance of context-specific AI models for supporting informed decisions by investors, tenants, and policymakers in dynamic housing markets.