The rapid advancement of artificial intelligence, exemplified by systems such as AlphaGo and large language models, has great potential to contribute to the development of human-like intelligence. However, fundamental differences exist between the underlying mechanisms of these systems and those of biological organisms. For instance, humans can achieve impressive performance with limited data and computing resources, while existing algorithms often require significant amounts of data and computing power for real-time operations. One of the reasons for this disparity is the human ability to plan in a model-based sense, making computational models that can capture human planning behavior valuable to bridge the gap between existing AI systems and human-like intelligence. This thesis explores the effectiveness of planning algorithms in modeling human behavior. Existing literature often overlooks timing information, and I develop a novel tree-based model that aims to capture both human action selection and human reaction times. The thesis also introduces a timing-sensitive goal recognition framework that incorporates timing information, and uses this framework to model human goal inference. My findings indicate that a Bayesian framework that incorporates a prior based on goal difficulty and a likelihood derived from an online planner accurately predicts human goal inference. This thesis underscores the promise of planning algorithms in mimicking human behavior and their utility in human-robot collaborations. More generally, it suggests that planning algorithms have an important role to play in advancing human-like intelligence.