Z.ai's GLM-5.2 Beats GPT-5.5, Intensifying AI Race

A Chinese AI startup, Z.ai, claims its new model, GLM-5.2, outperforms OpenAI’s GPT-5.5 on key benchmarks. This is intensifying the global artificial intelligence race. Z.ai, formerly Zhipu AI, unveiled GLM-5.2, highlighting its improved reasoning and software engineering capabilities. The company says it's designed for "long-horizon" tasks requiring sustained reasoning. On "Humanity’s Last Exam," GLM-5.2 scored 54.7, beating GPT-5.5's 52.2. It also showed gains on FrontierSWE, a benchmark for complex software engineering projects. What's interesting is GLM-5.2's ability to process large amounts of data, with a context window of up to one million tokens. This allows it to handle large codebases and long documents. Z.ai states the model can manage full-scale development workflows, from analyzing requirements to debugging. The bottom line is this launch highlights the growing competition between Chinese and American AI firms, especially as China's advancements narrow the gap in coding and applied engineering.