The turbulent state of the atmosphere has made it even more apparent the need to increase the accuracy of weather forecasts. Finding the best configuration to run the model with the highest accuracy requires testing these configurations. Also, since precipitation is one of the most difficult parameters, this study seeks to test WRF model configurations to better predict this variable. To do this, 9 different configurations with three nested domains with horizontal resoloution of 27, 9 and 3 km have been used for 24-hour forecasting on two different days in 2020 and 2019. The results showed that all configurations had a good ability to predict temperature and wind speed but in the case of precipitation, the configuration consisting of the Goddard micro-physics schema and the BMJ cumulus schema was more capable of predicting 24-hour winter precipitation on February 25th. Also in the case of spring precipitation The configuration with the combination of the Goddard sub-physics schema and the Grell cumulus schema was more capable.