This paper aims to develop Bayesian online change point detection (BOCD), a parametric change point detection method, into a nonparametric method to be able to detect change points in a free-distribution time series. Instead of using pre-defined exponential family distribution for predictive probability, we use kernel density estimation in which two possible options have been proposed. The first is manual constant bandwidth selection. This option provides a fast computation of KDE as it can pursue dynamic programming. Another option for maximum accuracy is a nonparametric bandwidth estimator. Additionally, to pursue the goal of fully nonparametric change point detection, the predefined hazard function in BOCD method is changed to be a nonparametric estimator. The performance of the proposed method was intensively evaluated with simulated and real-life data, and compared with other traditional methods. It was found that nonparametric BOCD gives a better solution in general cases as a consequence of the adaptive property of KDE. But this also comes with a drawback: it requires a sufficient amount of data to form a precise distribution curve to accurately detect change points.