Gas filled membrane absorption process (GFMA), or hollow fiber gas membranes process (HFGM) can replace two units operations (absorption and stripping) in a unique stage based on membrane contactors. But its application at industrial scale is still emerging: only cases of ammonia removal and concentration have been reported. This advantage has not been enough to scale it up to pilot prototypes for different applications, probably because (i) experimental tests have been carried out with modules that do not exist at industrial scale, and the selection of the mass transfer correlation for different purposes lacks rigour with respect to the original conditions of the correlation, and (ii) industrial modules configurations focused only on gas absorption applications. The first problem limits the experimental results for scientific purposes only, making these results un-reproducible for the design of a pilot or industrial plant. Furthermore, the selection of experimental modules has generally been established to validate an application, limiting the search for an optimal performance. The high quantity of mass transfer correlation also complicates the selection to design or scaling-up purposes. This work includes then a review of the first problem, by conducting an evaluation of mass transfer performance of different membranes modules configuration proposed in literature. The aim is to identify the best experimental configurations for a GFMA process that will lead a further scaling up. This evaluation was carried out using a phenomenological model of GFMA process developed for an application of cyanide recovery in the gold mining industry. The best performance was achieved in membranes modules having cross flow with center baffle and fully baffled.