This research focuses on benchmarking heterogeneous treatment effect (HTE) estimation algorithms in networked environments to enhance our understanding of causal relationships. By evaluating models such as X-Learner, T-Learner, and Causal Forest across synthetic, semi-synthetic, and real-world datasets, this work addresses the challenges posed by confounding, mediation, and interference in social networks. Through rigorous dataset generation, model tuning, and performance evaluation using metrics like ATE error and PEHE, the study highlights the strengths and limitations of these algorithms. Key findings demonstrate the variability in model performance under different conditions and underscore the need for context-aware model selection. This comprehensive benchmarking framework aims to inform future developments in causal inference methodologies, advancing robust and scalable solutions for complex network environments.