Z.ai releases GLM-5.2 on June 16, claiming Claude Opus 4.8-level results without Nvidia chips
Beijing-based AI lab Z.ai released its GLM-5.2 model on June 16, saying it delivers performance close to Claude Opus 4.8 across several open benchmarks while being trained entirely on Huawei Ascend chips, with no Nvidia hardware. The company has leaned harder into domestic computing after being added to the U.S. Entity List in January 2025. Its shares rose 90% over the past week to a record high.