CAMS provides global solar radiation estimates for clear-sky (McClear model) and all-sky (Heliosat-4 method) conditions, the latter based on MSG satellite information. A performance assessment of these estimates (with site-adaptation and spatial smoothing) is done, using hourly data from 10 sites in the Pampa Humeda region of South America. Two locally adjusted Cloud Index Models (CIM) using GOES-East satellite information are also evaluated. One of them (CIM-ESRA) is based on the ESRA clear-sky model and the other (CIM-McClear) on the McClear clear-sky model. Under clear-sky conditions, the site-adapted McClear is found to perform best with a relative root mean square deviation (rRMSD) of 2.8%. However, in the presence of clouds in the real atmosphere, the model tends to provide lower clear-sky estimates than the ESRA model which, in our implementation, is only sensitive to average atmospheric trends. Under all-sky conditions, both CIMs show a small but consistent underestimation of - 1.1% in the region and perform significantly better than the site-adapted Heliosat-4, with rRMSDs of 12.1% (CIM-McClear), 12.5% (CIM-ESRA) and 16.8% (site-adapted Heliosat-4). This performance difference is not a statement about the relative quality of the models, since it can be explained by the difference in satellite view angle (significantly higher for the MSG satellite than for the GOES-East satellite). The performance downgrade due to using MSG satellite images out of their recommended area is quantified. Both CIMs, based on using GOES-East imagery, provide accurate solar irradiation estimates over this region and can be extended to other areas of Latin America.