Precise land use and land cover (LULC) classification is challenging because of urban heterogeneity, which is essential for sustainable urban planning and scientific research. Hence, there’s a growing need for accurate LULC maps, thus comparing the various image classification models is necessary to choose the best one. Therefore, the present study aims to compare the performance of various classification models. Currently, Google Earth Engine (GEE) is the most innovative open-source platform worldwide for advanced geospatial big data research. In this research, Random Forest (RF) and Support Vector Machine (SVM) machine-learning algorithms were applied to this cloud platform for LULC classification and then compared with the traditional Maximum Likelihood classification (MLC) method for the Kota city. Landsat-8 datasets were used for 2023 and analysis was done in ArcGIS 10.6 software. Accuracy and validation assessments were used to evaluate each of the three models’ performance. The results showed that the Kappa coefficients for RF, SVM, and MLC are 0.97, 0.95, and 0.94, respectively, while the overall accuracy for these three models are 98.68%, 97.00%, and 96.25%, respectively. The current study demonstrates that RF outperformed SVM and MLC in this classification and comparison. According to the present research study, GEE effectively processed satellite imagery and created four classes of LULC maps of the research region: built-up area, vegetation, waterbody, and barren land. The best accuracy results of RF LULC classification can be used for further analysis related to sustainable development.