Pakistan is one of the countries heavily dependent on hydrocarbon fuel for energy production which is causing a severe climate change; however, wind energy seems to be a long-term solution. Various statistical distributions have been used to draw the analysis of wind data, but the selection of an optimum method has been a challenge. This work is an assessment of wind power potential of a site located near the southern coast of Pakistan. Data collected in two years is analyzed at four heights using three variations of Weibull parameters. Weibull parameters computed through the proposed mean bias error based artificial intelligence grey wolf optimization were compared with the computations through Rayleigh and Justus's empirical numerical methods. Root mean square error, determination of coefficient, and mean bias error, are computed to validate and compare the computed results. The wind characteristics like most probable and Maximum energy-carrying wind are found to be in excellent compatibility with most of the wind turbines that could be used on the site. The results obtained along with a brief cost analysis, for eight selected wind turbine systems, show that the considered site is suitable for the production of a wind power project. (C) 2020 Elsevier Ltd. All rights reserved.