This paper reviews the application of Physics-informed Neural Networks (PINNs) in fluid mechanics, highlighting their ability to integrate data and mathematical models seamlessly. PINNs are particularly useful for solving inverse problems, which are often challenging and costly with traditional numerical methods. The authors demonstrate the effectiveness of PINNs in three-dimensional wake flows, supersonic flows, and biomedical flows. They show that PINNs can accurately infer flow fields from limited data, making them a promising tool for complex flow problems. The paper also discusses recent advancements in PINNs, including domain decomposition and uncertainty quantification, and provides detailed case studies to illustrate their performance. Overall, PINNs offer a complementary approach to existing numerical methods, particularly for realistic fluid flows with noisy or incomplete data.This paper reviews the application of Physics-informed Neural Networks (PINNs) in fluid mechanics, highlighting their ability to integrate data and mathematical models seamlessly. PINNs are particularly useful for solving inverse problems, which are often challenging and costly with traditional numerical methods. The authors demonstrate the effectiveness of PINNs in three-dimensional wake flows, supersonic flows, and biomedical flows. They show that PINNs can accurately infer flow fields from limited data, making them a promising tool for complex flow problems. The paper also discusses recent advancements in PINNs, including domain decomposition and uncertainty quantification, and provides detailed case studies to illustrate their performance. Overall, PINNs offer a complementary approach to existing numerical methods, particularly for realistic fluid flows with noisy or incomplete data.