The Dee Investigation Simulation Program for Regulating Network (DISPRIN) model is a type of lumped model. This model has 25 parameters whose values are continuous so that it is difficult to apply to solve practical problems. This study aims to improve the performance of DISPRIN so that it is effective and applicable to generate historical discharge data series in a watershed. The combination of the simulation equation system from the DISPRIN model with the parameter optimization method based on the metaheuristic method is expected to produce a new model that is able to carry out the calibration process automatically so that the model becomes easy to apply. The metaheuristic methods involved are: Differential Evolution (DE) Algorithm, Particle Swam Optimization (PSO), synthesis of chaotic search-opposition based learning-differential evolution-quantum mechanism (CODEQ) algorithm, and Shuffled Complex Evolution (SCE). The new models produced are then called the DISPRIN-de, DISPRIN-pso, DISPRIN-sce, and DISPRIN codeq models. All models were tested in Lesti watershed (314.19 Km2), Malang Regency, East Java Province, Indonesia. The model calibration stage using hydroclimatology data from 2006 to 2014 showed that all models had an accuracy level equivalent to NSE ranging from 0.892 to 0.931, and the model validation stageusing hydroclimatology data from 2014 to 2020 produced NSE values ranging from 0.918 to 0.928. The discharge distribution curve involving all generated discharges showed that the DISPRIN-codeq model was more accurate than the other three models which tended to overestimate high flow events.