The urgent need for environmentally friendly alternative fuels arises from concerns over fossil fuel depletion and the harmful effects of exhaust gas emissions. One promising solution involves the use of biodiesel blended with natural additives like essential oils to improve combustion efficiency in diesel engines. However, conventional engine performance testing is often time-consuming and repetitive, requiring a more efficient approach such as predictive modeling using Machine Learning (ML). This study investigates the effect of biodiesel and essential oil blend ratios on engine performance and develops predictive models to estimate engine power and torque. A total of 1045 experimental data points were collected from tests using varying compositions of biosolar, dexlite, and essential oil additives at different RPMs. The research applied a quantitative method, utilizing Linear Regression and Support Vector Machine (SVM) algorithms implemented in RStudio. Model accuracy was evaluated using MAE, RMSE, and R² metrics. The results indicate that variations in biodiesel blend ratios and essential oil additives had no statistically significant effect on engine performance. However, ML-based predictive modeling proved highly effective. The SVM model achieved superior accuracy, with R² values of 0.993 for power and 0.973 for torque. In contrast, the Linear Regression model yielded much lower R² values—0.671 for power and 0.093 for torque. These findings demonstrate the potential of Machine Learning, particularly SVM, as a reliable tool for predicting diesel engine performance using biodiesel and additive blends, offering a faster and more accurate alternative to conventional testing methods.