An era has started in which gas and dust can be observed independently in protoplanetary disks, thanks to the recent surveys with ALMA. The first near-complete high-resolution disk survey in both dust and gas in a single star-forming region has been carried out in Lupus, finding surprisingly low gas/dust ratios. The goal of this work is to fully exploit CO isotopologues observations in Lupus, comparing them with physical-chemical model results, in order to obtain gas masses for a large number of disks. We have employed physical-chemical models to analyze continuum and CO isotopologues observations of Lupus disks, including isotope-selective processes and freeze-out. Employing also the ALMA 13CO-only detections, disk gas masses have been calculated for a total of 34 sources, expanding the sample of 10 disks studied by Ansdell et al. (2016), where also C18O was detected. We confirm that overall gas-masses are very low, often smaller than 1 $MJ$, if volatile carbon is not depleted. Accordingly, global gas/dust ratios predominantly between 1 and 10. Low CO-based gas masses and gas/dust ratios may indicate rapid loss of gas, or alternatively chemical evolution, e.g. via sequestering of carbon from CO to more complex molecules, or carbon locked up in larger bodies. Current ALMA observations cannot distinguish between these two hypotheses. We have simulated both scenarios, but chemical model results do not allow us to rule out one of the two. Assuming that all Lupus disks have evolved mainly due to viscous processes over the past few Myr, the observed correlation between the current mass accretion rate and dust mass found by Manara et al. (2016) implies a constant gas-to-dust ratio, which is close to 100 based on the observed $Mdisk/M\dot{}acc$ ratio. This in turn points to a scenario in which carbon depletion is responsible for the low CO isotopologue line luminosities.