CORL provides high-quality, easy-to-follow implementations of state-of-the-art offline reinforcement learning algorithms including AWAC, BC, CQL, DT, EDAC, IQL, SAC-N, TD3+BC, and others. Each algorithm is implemented as a single file for research clarity, backed by a benchmarking system and Weights & Biases integration for experiment tracking. The library uses d4rl and gym interfaces for environment compatibility.