Mapping of mangrove extent is considered one of the important aspects of management and monitoring of the mangrove ecosystem as they play a vital role in protecting the areas along the coast. Mangroves have a unique habitat consisting of various marine and land organisms that are indigenous to the region and play a key role in maintaining the balance in the ecosystem. Satellite imageries and advanced classification techniques of machine learning provide a better chance of mapping and delineating these inaccessible areas. The Google Earth Engine (GEE) is one-of-a-kind cloud-based tool that allows the integration of satellite imagery analysis by implementing advanced machine learning algorithms on an online platform. In this study, two machine learning classifiers such as Random Forest (RF) and Classification and Regression Trees (CART) are compared for mapping mangrove and non-mangrove classes using the GEE platform. The Sentinel-2 imagery from January to March 2018 and 2022 over the creeks along the coast of Maharashtra was used in the study for delineating the mangrove cover. The two algorithms were tested for suitability and accuracy for mapping mangrove vegetation. The accuracy assessment in mapping the mangrove extent was 89.7% and 91.05% for 2018 and 89% and 92.04% for 2022 from the RF and CART algorithms, respectively. The limitation of the study using these two methods is that the parameters used are the default parameters available in GEE. All classifiers in GEE used these training areas to classify the image into specified land cover classes. The classification results demonstrate that all classifiers classified the mangrove pixels. The map results were compared to the visually interpreted mangrove map of the study area as references to assess the correctness of the map produced. In this comparison, we discovered that RF greatly exceeded the classification results of CART. This work demonstrates the capabilities of GEE for mapping mangrove extent.