In this article, we present probabilistic energy models for hybrid energy harvesting (HEH) Internet of Things (IoT) nodes. We aim to model the energy obtained from multiple energy sources in a single form with mixture models and to obtain energy characteristics at every stage of HEH. Multiple energy sources are classified into energy source clusters (ESCs), and harvested energies at the output of energy harvesting modules are modeled by using Gaussian mixture models (GMMs). The joint distributions of harvested energies with GMMs are expressed as the sum of the joint densities produced by the possible combinations of component densities. A generalized expression is presented for both independent and correlated ESCs. The distributions of energy levels are obtained based on the joint densities at the output of the energy combiner unit as well as at the energy storage unit. This approach provides a single form and an accurate probabilistic model for each cluster including multiple sources, taking the randomness of ambient energy into account in HEH. Depending on the energy requirements, the characteristics of an energy storage unit can be determined for IoT applications.