SocialED is a comprehensive, open-source Python library for social event detection (SED), integrating 19 algorithms and 15 datasets under a unified, modular API. It supports diverse preprocessing techniques (e.g., graph construction, tokenization) and provides standardized interfaces for training and inference. Designed for extensibility and efficiency, SocialED is compatible with popular deep learning frameworks and runs smoothly on both CPU and GPU. The library emphasizes code quality through unit testing, CI, and coverage checks. It is available at https://github.com/RingBDStack/SocialED and installable via PyPI. It has gained significant attention, with more than 580 stars on GitHub and over 18,000 PyPI downloads.