Lithium-ion battery state-of-health (SOH) is critical in measuring the degradation and safe operation conditions. Proper prediction methods are crucial for obtaining accurate SOH values. This paper applies the Gaussian Mixture Model (GMM) and Gaussian Mixture Regression (GMR) to the extracted features to predict battery SOH in a probabilistic approach. The probabilistic approach represents a dataset as a combination of multiple Gaussian distributions, allowing it to capture complex data distributions. The features are extracted from the Constant Current (CC) charging stage. Feature 1 is the maximum point of the rate of charge increase from 3.5 V to 4.2 V, while feature 2 is the maximum capacity increase from 3.7 V to 4.2 V, both with a step of 0.1 V. The dataset used is from the University of Maryland and contains different charging profiles. The root mean square error (RMSE) is less than 1.5%, and the R2 values are over 99% when applying the GMR model. The preliminary study on the application of GMM and GMR shows a reliable performance in SOH prediction, and further studies based on GMM and GMR can be done in the future.