This study aims to describe the difference in student learning outcomes using the numbered head together (NHT) learning model and the problem based learning model. the type of research used is quasi experiment research with non-equivalent control group design using a quantitative approach. The research was carried out at State Vocational High School 1 Kediri Building Modeling and Information Design Expertise Program and took place in the odd semester of the 2025/2026 school year. The sampling technique used for this study was with a purposive sampling technique of 36 students as an experimental group and 33 students as a control group. The instrument used in this study is a test of 20 multiple-choice questions in measuring student learning outcomes. The data analysis technique in the study used a different test (t-test). The results of the study showed a significant difference in student learning outcomes using the Numbered Heads Together model (0.000) compared to the Problem-Based Learning model. The NHT model prioritizes collaboration and active participation, which has been proven to improve student learning outcomes. The NHT model prioritizes collaboration and active participation, proven to improve student understanding. This is shown by the interaction between students can strengthen their understanding through explaining to each other. In addition, the NHT model also encourages students to be more confident in expressing their opinions. Thus, the results of this study confirm the interactive and collaborative NHT model in improving student learning outcomes in the future.