Current satellite-based systems for measuring soil moisture don't have high enough resolution for many important uses, like managing farm irrigation, predicting floods and droughts, or modeling how water moves through the environment. To solve this problem, we created a new global soil moisture dataset called GSSM-10, which provides data at a much finer 10-meter resolution. We used information from several types of satellites—including radar, optical sensors, thermal imagery, and elevation data—and combined them using advanced machine learning models. These models were trained using real soil moisture measurements from around the world. Our final model performed very well, both during testing and on completely new data. It can accurately estimate how much moisture is in the top layer of soil. We've also built an easy-to-use online platform where anyone can explore, download, and use the data. GSSM-10 offers a powerful new tool for farmers, scientists, and decision-makers who need detailed and reliable soil moisture information.