China's Z.ai says new model nears Anthropic's Mythos 5 in cyber-defence tests

Chinese AI startup Z.ai said its open-source GLM-5.3 model has neared Anthropic’s restricted Mythos 5 at identifying software vulnerabilities, signaling narrowing performance gaps with a leading US model. The update boosts the credibility of a China-based challenger that is gaining traction among Western developers, a modest positive for sentiment around AI competition and open-source capability.

Analysis

Chinese AI startup Z.ai said its open-source GLM-5.3 model has neared Anthropic’s restricted Mythos 5 at identifying software vulnerabilities, signaling narrowing performance gaps with a leading US model. The update boosts the credibility of a China-based challenger that is gaining traction among Western developers, a modest positive for sentiment around AI competition and open-source capability.