The DexonomySim dataset, specifically tailored for robotic dexterous grasping simulations, has been officially released as an open-source resource. Comprising over 9.5 million high-quality grasping poses, this extensive dataset encompasses more than 10,000 objects and 31 commonly utilized grip types. It stands as the largest open-source, high-quality synthetic dataset available for dexterous hand manipulation in the realm of embodied intelligence.