Zhipu AI's GLM-4.1V-Thinking is a cutting-edge open-source vision-language model that excels in complex reasoning tasks. Despite its 9 billion parameters, it matches or surpasses the performance of larger models, including the 72 billion-parameter Qwen-2.5-VL, across multiple benchmarks. This model supports diverse inputs such as images, videos, and documents, and offers a 64k context length, enabling it to handle extensive information effectively. A demo and API are available for users to explore its capabilities firsthand.