Zhipu announced the open-source new generation flagship model GLM-4.7.

Today, Zhilu released and open-sourced its new flagship model GLM-4.7. This version achieved the best performance among open-source models in multiple mainstream public benchmarks. In the authoritative coding evaluation system Code Arena, which has millions of global users participating in blind tests, GLM-4.7 ranks first in open source and domestic models, surpassing GPT-5.2. It is reported that GLM-4.7 enhances coding ability, long-term task planning, and tool collaboration in coding scenarios, in addition to improving overall performance in chat, writing, and role-playing.