The TianGong large model is a massive language model developed by Kunlun Tech, boasting a scale of over a billion parameters. It has recently demonstrated exceptional performance on multiple evaluation datasets. The model surpasses both domestic and international large language models in comprehensive assessments, particularly excelling in language comprehension and logical reasoning. This makes the TianGong model a top contender among closed-source models for Chinese, holding significant implications for the field of Chinese natural language processing. However, there are still some limitations in the evaluation results that require further research and improvement.