The ability of nuclear mass models to describe experimentally known nuclear masses is studied. Twelve frequently used models of various kinds are considered. From the selected models, eleven are of global character and one is a local model specially adapted to describe heavy nuclei. To decrease the number of nuclei over which the accuracy is averaged, the global region ($Z,N\ge 8$) is divided into four subregions, in which the accuracy is studied separately. Finally, to reach the best precision, the accuracy is investigated without any averaging for each nucleus separately. The results are presented in the form of colored maps, large enough to be easily and accurately read.

Besides the accuracy of the studied models, their predictive power is also considered.

It is found that the accuracy of the description of the mass strongly depends on the nuclear-mass model and on the region of nuclei to which the model is applied. The best accuracy is obtained by the two recent Chinese models WS3+ and WS4+. Generally, no clear, strong correlation between the accuracy of the description of already known masses by a given model and its predictive power for new masses is observed. However, such a correlation is found for separate models and in separate regions of nuclei, more often for the macroscopic–microscopic models than for the other approaches.

研究了核质量模型描述实验已知核质量的能力。考虑了十二种常用的各种模型。从选定的模型中，有11个具有全局特征，一个是局部模型，特别适合描述重核。为了减少平均精度的核数目，全局区域（$ž，ñ\ge 8$）分为四个子区域，在其中分别研究准确性。最后，为了达到最佳精度，对精度进行了研究，而没有对每个原子核分别求平均值。结果以彩色地图的形式显示，其大小足以轻松准确地读取。

除了研究模型的准确性外，还考虑了它们的预测能力。

发现质量描述的准确性很大程度上取决于核质模型和应用该模型的核区域。最近的两个中文模型WS3 +和WS4 +获得了最佳的准确性。通常，在给定模型的已知质量描述的准确性与其对新质量的预测能力之间，没有观察到明显的强相关性。但是，在单独的模型和原子核的不同区域中发现了这种相关性，与其他方法相比，更常见的是宏观-微观模型。