Scientists evaluated different climate models to see how well they predict temperature and rainfall in China. They used a new method to compare the models' performance in simulating past climate and predicting future changes. Some models did better than others, with CMCC, IPSL-CM5A-MR, and BCC-CSM1.1M performing well globally. Models from MPI and NCAR-CESM1 were closest to the average predictions. Higher resolution improved historical climate simulations but didn't affect future climate predictions much. No single model was best at both historical and future climate predictions, as the "best" model depends on the specific use. This study helps groups choose climate models for their own research.