The invention discloses a model training method and device, computing equipment and a storage medium, and belongs to the technical field of computers. The method comprises the steps of obtaining corpus data used for performing voice control on smart home equipment; determining whether the corpus data meets a model training triggering condition or not; and when the corpus data meets the model training triggering condition, performing model training on a voice control model according to the corpus data to obtain an updated voice control model. Model training is carried out only when corpus datais accumulated to one degree, so that the number of times of model training can be reduced to a certain extent, the consumption of training resources is reduced, the model training resources can be saved, and meanwhile, the efficiency of model training is improved by reducing the number of times of model training.