Air pollution is a major problem in cities across the world, especially in developing countries like India. This study aimed to compare the sentiment analysis of air quality perception in two major metro cities, Delhi and Mumbai using natural language processing techniques. The hypothesis was that the sentiment expressed on social media platforms like Twitter and news articles can provide insights into the public perception of air pollution. The process involved use of machine learning models such as support vector machines and naive bayes, along with tf-idf and bag-of-words algorithms, to analyze a data set of 0.1 million tweets related to air quality collected in the time period of Jan 2022 - Dec 2022. The findings of this study have important implications for policymakers and stakeholders concerned with improving air quality in India. The strong correlation observed between the air quality index and sentiment of tweets suggests that social media can be an effective tool for monitoring air quality and understanding public perception of air pollution. The results also highlight the need for targeted interventions to address the problem of air pollution in these cities. For instance, the negative sentiment expressed in tweets related to air quality in Delhi indicates the need for urgent measures to mitigate the sources of pollution in the city. In contrast, the less negative sentiment expressed in Mumbai suggests that the city may have implemented some successful interventions that could be studied and potentially replicated in other cities. This study only focuses on two of the major metro cities in India and may not be representative of the entire country. However, by utilizing sentiment analysis of social media and news articles, policymakers and stakeholders can gain valuable insights into public perception of air pollution, which can inform policy decisions and public awareness campaigns aimed at improving air quality in major metro cities.