Zhipu unveiled the groundbreaking open-source visual reasoning model, GLM-4.5V, yesterday. This model impressively features a total of 106 billion parameters, with 12 billion active parameters. Despite being categorized as a 100B-level model, GLM-4.5V surpasses this threshold, positioning it as the top-performing model in its class. The model has been concurrently open-sourced on both the ModelScope community and Hugging Face, further enhancing its accessibility and impact.