This paper introduces a hybrid framework, HiSim, for simulating social media user behavior in large-scale social movements. The framework categorizes users into two types: core users, driven by Large Language Models (LLMs), and ordinary users, modeled by deductive agent-based models (ABMs). A Twitter-like environment is constructed to replicate user responses to trigger events. The authors develop a benchmark, SoMoSiMu-Bench, to evaluate the simulation's effectiveness using real-world datasets. Experimental results demonstrate the framework's effectiveness and flexibility in simulating social movement dynamics. The contributions include a novel hybrid simulation framework, a tailored simulator for online social movements, and a comprehensive benchmark for evaluation. The paper also discusses the formalization of public opinion dynamics simulation, the hybrid framework for social media simulation, and the evaluation of the benchmark. The scalability and ethical considerations of the method are also addressed.This paper introduces a hybrid framework, HiSim, for simulating social media user behavior in large-scale social movements. The framework categorizes users into two types: core users, driven by Large Language Models (LLMs), and ordinary users, modeled by deductive agent-based models (ABMs). A Twitter-like environment is constructed to replicate user responses to trigger events. The authors develop a benchmark, SoMoSiMu-Bench, to evaluate the simulation's effectiveness using real-world datasets. Experimental results demonstrate the framework's effectiveness and flexibility in simulating social movement dynamics. The contributions include a novel hybrid simulation framework, a tailored simulator for online social movements, and a comprehensive benchmark for evaluation. The paper also discusses the formalization of public opinion dynamics simulation, the hybrid framework for social media simulation, and the evaluation of the benchmark. The scalability and ethical considerations of the method are also addressed.