As agriculture remains a primary source of income in India, the rising instances of farmer distress, often associated with crop failure and low yields, emphasize the immediate need to promote agricultural sustainability. This challenge extends its implications to both the nation's economy and the global agricultural sector, becoming a significant concern for countries worldwide. The key issue revolves around the precise prediction of crop yields under diverse climatic conditions in India, considering various attributes. The overarching objective is to equip farmers with essential tools for informed decision-making, enabling them to choose the most suitable crops and optimize fertilizer usage. In the proposed system, Machine Learning (ML) model is used to address a variety of agricultural challenges, including soil classification and crop recommendation. Addressing these challenges, a proposed model aims to categorize and forecast suitable crops based on soil nutritional levels, employing diverse ML models termed as K-Nearest Neighbors (KNN), Logistic Regression (LR), Support Vector Machine (SVM), Stochastic Gradient Boosting (SG), and XG-Boost (XGB). The research dataset encompasses seven distinct soil parameters characterized by seven nutrient attributes. Training and testing of the models are conducted using 80 and 20% of the dataset, respectively. Notably, the outcomes reveal that XG-Boost surpasses other models, achieving an accuracy of 98.64%, precision of 98.49%, recall of 98.76%, and an F-score value of 98.59%.