Large Indian cardamom (Amomum Subulatum Roxb.), is a very popular and well known spice crop for its aromatic properties, mainly presence of many volatile bio-active compounds like 1,8-cineole, alpha pinene and beta pinene. These compounds play a crucial indicator for cardamom quality, influencing consumer priorities and commercial appeal. There are several conventional chemical analysis techniques are available, which are reliable but its involved elaborate procedures and time consuming. In this research, we examined a rapid, non-destructive method for prediction of 1,8-cineole concentration in large cardamom samples using near infrared (NIR) spectroscopy integrated with different machine learning models. Large cardamom samples were collected from six agro-climatically diverse regions across India to capture a wide range of variability. The NIR spectral data underwent preprocessing steps including z-score normalization, jitter-based augmentation and selective filtering to enhance uniformity and to reduce the noise. Different regression algorithms namely Partial Least Squares Regression (PLSR), Decision Trees (DT), Random Forest (RF), Support Vector Regression (SVR) and Extreme Gradient Boosting (XGBoost) have been applied on the preprocessed dataset for the prediction of 1,8-cineole concentration. The predictive model of 1,8-cineole concentration were evaluated using various performance metrics such as coefficient of determination (R2), root mean squared error (RMSE), mean absolute error (MAE) and mean absolute percentage error (MAPE). The resultant value of performance metrics indicate strong predictive performance.