Computational tools for searching molecular databases accelerate lead identification in drug discovery. In this work, we introduce qMol, an online platform designed to enable the search for accessible molecules using a fragment-based reduced graph representation, wherein molecules are represented as chemically meaningful reduced graphs with fragments as nodes and their connections as edges. qMol enables flexible molecular searches by allowing users to retrieve exact or varied reduced graphs and apply fragment-level constraints, providing precise control over global topology and local chemical features to support analogue identification in drug discovery. We hope that qMol will be a useful resource for medicinal researchers to efficiently explore chemical databases. qMol is publicly available and can be accessed at https://qmol.xundrug.cn.