This study explores the use of machine learning techniques (ML) to evaluate ultimate tensile strength (UTS) for friction stir welding (FSW). The exploration of FSW, a solid-state welding technology, is driven by its better structural integrity, reduced defects, and improved mechanical qualities. Reducing recurrent experimentation and UTS testing is the main issue that is addressed, especially for AA6061 metal. The main goal is to use ML algorithms to forecast UTS values which saves time and resources. Several machine learning techniques are used in this process, in this paper Random Forests, Decision Trees, Gradient Boosting, and KNN are being implemented. The results reveal that the Gradient Boosting, Decision Trees, and Random Forests algorithms can forecast UTS values accurately, indicating their potential use in all FSW procedures. This discovery creates opportunities for additional study and use. The importance of ML in FSW technique optimization and its wider implications for industrial and structural applications are emphasized in the conclusion. Potential future developments in the field of welding technology include increasing the model’s complexity, adding more parameters, providing real-time monitoring and control, and expanding the model’s transferability to other materials.