Beijing-based Zhipu AI released GLM-5.2 on June13. The AI system is an open-weight model, meaning that its core components are publicly released, allowing users to modify it as they wish with limited restrictions.
On Friday, The Wall Street Journal reported that GLM-5.2 capabilities matched that of Mythos in bug-finding. In some capabilities, it outshone Anthropic’s Claude Opus 4.8 model, the report said. Crowd-sourced ranking platform Code Arena WebDev Leaderboard listed GLM-5.2 second to Claude Fable-5 in terms of capabilities. The model has also been categorized as the “lowest cost per task among models at its intelligence level.”
Zhipu claims the latest model has “stronger coding capabilities with multiple thinking effort levels to balance performance and latency.” With the AI system being equipped with 1 million token context, it can support “large-scale implementation, automated research, performance optimization, and complex debugging,” the company said.
The announcement from the Chinese company comes amid mounting concerns, especially in the US over advanced AI model’s malicious cyber capabilities. This month, OpenAI released its latest model GPT5.6 with limited access, citing a U.S. government request. Similarly, Anthropic also imposed access limitations on its models, Fable-5 and Mythos-5, following an US government intervention.
The release of this GLM-5.2 will raise questions about whether access restrictions on Western models can hold when equivalent tools are freely available from China.
The concerns from the U.S. government comes as it continues to assess the disruptive capabilities of latest AI models. Security experts believe nation-state hackers and other malicious entities could use the models for mass vulnerability scanning and its weaponizing, leading some to dub the potential scenario as “bugmageddon.”
The release of the latest model is also fast closing the AI capabilities race gap between China and the US. In 2025, the launch of generative AI R1 model by Chinese AI company DeepSeek wiped out $1 trillion from the US tech stock markets.
Although the release of GLM-5.2 did not have similar negative impact, the latest model has certainly raised some concerns.
“China is making sure that the gap becomes smaller and smaller over time,” said Lior Div, chief executive officer of the cybersecurity company 7AI told The Wall Street Journal.
