The background of this research is the poor literacy ability of Indonesian students based on data from the TIMSS (Trend in International Mathematics and Science Study) and the results of the PISA test. (Programme for International Student Assessment). The aim of the study was to find out how students' numeration literacy was improved using the cooperative type teams games tournament (TGT) model and how the cooperative type TGT model affected the numeration skills of elementary school students. This kind of research is an experimental quasi-sample. In this study, there are as many as 20 students from the experimental class and 20 students from the control class. Students in the experiment class who are treated with the TGT model and those in the those in the control class only receive treatment with the conventional model. The test instrument in this study is a description-shaped test that measures the numerical literacy ability of the student being studied. The result of this study is seen in the average score of students before receiving treatment in the experimental class of 49.00. Then there was an increase in student posttest scores after obtaining learning treatment using the cooperative type TGT model to 86.00. And the calculation of the regression equation test showed that the data from the pretest and posttest obtained a value of  = . From the determination coefficient on the results of the regression analysis of the experimental class, the influence of the model used was 61.8%. It can be concluded that the numeration literacy of students in the experiment class improved after being treated with the Cooperative Type TGT model, and there was a major influence on the TGT model on the numerational literacy ability of the students in the experimental classes.