MARKLLM is an open-source toolkit for LLM watermarking that provides a unified and extensible framework for implementing watermarking algorithms, along with user-friendly interfaces and visualization tools. It supports nine algorithms from two families: KGW and Christ. The toolkit includes 12 evaluation tools across three perspectives—detectability, robustness, and text quality impact—and two automated evaluation pipelines. MARKLLM also offers visualization solutions to help users understand the mechanisms of different algorithms. The toolkit is designed with a modular architecture, enabling easy integration of new algorithms and visualization techniques. It provides a comprehensive set of resources, including a Python package, Jupyter notebook demo, and detailed documentation. MARKLLM has gained significant attention from researchers and developers, with active contributions to its development. The toolkit supports various evaluation metrics, attacks, and tasks, enabling researchers to conduct thorough evaluations of watermarking algorithms. MARKLLM facilitates the comparison of different algorithms in terms of detectability, robustness, and text quality impact. The toolkit also includes user examples and scripts for evaluating watermarking algorithms, making it accessible for researchers to conduct experiments and analyze results. Overall, MARKLLM serves as a valuable resource for advancing research and applications in LLM watermarking technology.MARKLLM is an open-source toolkit for LLM watermarking that provides a unified and extensible framework for implementing watermarking algorithms, along with user-friendly interfaces and visualization tools. It supports nine algorithms from two families: KGW and Christ. The toolkit includes 12 evaluation tools across three perspectives—detectability, robustness, and text quality impact—and two automated evaluation pipelines. MARKLLM also offers visualization solutions to help users understand the mechanisms of different algorithms. The toolkit is designed with a modular architecture, enabling easy integration of new algorithms and visualization techniques. It provides a comprehensive set of resources, including a Python package, Jupyter notebook demo, and detailed documentation. MARKLLM has gained significant attention from researchers and developers, with active contributions to its development. The toolkit supports various evaluation metrics, attacks, and tasks, enabling researchers to conduct thorough evaluations of watermarking algorithms. MARKLLM facilitates the comparison of different algorithms in terms of detectability, robustness, and text quality impact. The toolkit also includes user examples and scripts for evaluating watermarking algorithms, making it accessible for researchers to conduct experiments and analyze results. Overall, MARKLLM serves as a valuable resource for advancing research and applications in LLM watermarking technology.