Precipitating and convective core cloud hydrometeors and their radiative effects are generally ignored in global climate models (GCMs) such as those used in IPCC 4th and 5th Assessment Reports. However, remote sensors such as CloudSat and CERES are sensitive to all particles, including precipitating and floating cloud hydrometeors and their radiation. We examine GCMs within CMIP3 and CMIP5 with respect to their representation of ice and liquid water with exclusion of precipitation and convective cloud profiles that is provided with each CloudSat and CALIPSO combined retrieval. Surface sensible and latent heat fluxes from reanalyses, as well as CERES and SRB radiation data, are used for model radiation and global energy budget evaluations. Model evaluation also includes the cloud and radiation satellites simulator output available in IPCC 5th assessments. Significant disparity is found in the magnitudes, spatial distributions and correlations against estimates among the CMIP3 and CMIP5 models. The global energy budget is well in-balance at the TOA implying that models might be overly tuned by only a subset of the observations. The largest bias in the global surface energy budget is from the downward shortwave flux resulting from excessive net downward surface fluxes over convectively active regions compared to CERES/SRB data. This may be caused by not treating the interaction of precipitation with radiation in the models.