We summarize PCMDI's leading international role to establish capabilities for the systematic benchmarking of all global climate models. Building on its ongoing responsibilities for CMIP, PCMDI is now also engaged in a parallel effort (called obs4MIPs) to facilitate model evaluation by promoting production of observational datasets that are structured and formatted consistent with model output. PCMDI is also building a capability to provide baseline metrics of model performance on a routine basis for the core CMIP experiments, referred to as the CMIP Diagnostic, Evaluation, and Characterization of Klima (DECK) experiments. This metrics package will be made available to climate modeling groups who may use it during model development to help in identifying strengths and weaknesses in their model, relative to the ensemble of CMIP models. Prior to its general release, the package is being tested by three modeling groups.