Rapid urbanization demands solutions for sustainable healthy smart cities. It aims to make cities safe, environmentally friendly, equitable, and resilient. One of the critical problems in cities is air pollution. Monitoring air quality is essential for ensuring public health. The established remote air quality monitoring stations in smart cities produce a large volume of sensor data on air pollutants. Intelligent systems to analyze air pollution to build a sustainable solution for smart cities are of the utmost importance. The current study reports the applicability of several machine learning (ML) models for air pollutants analysis in major smart cities in India. The work focuses on three major tasks: 1) multiclass categorization of air quality using a variety of classifiers, 2) Air Quality Index (AQI) prediction by employing different regression models, and 3) comparative study of empirical results of different classifiers and regressors used in air quality analysis of smart cities. The research employs five classifiers and five regressors to analyze real-time sensor data of air pollutants from major five Indian metro cities from 2021 to 2023. The performance of these learning models is evaluated on a variety of metrics.