During recent years stochastic algorithms have deserved much attention from the computational biology research communities. In this paper we derive a hybrid version of the formerly known Metabolic Algorithm that is enriched with stochastic features, whose impact on the dynamics of the system is especially prominent when the amount of metabolite becomes smaller. This hybrid procedure represents a first attempt to let the Metabolic Algorithm deal with low concentrations of substances according to a non-deterministic policy.