GLM-5.3 superou ligeiramente o Mythos 5 em um teste de segurança cibernética

BEIJING: Chinese AI startup Z.ai said on Friday (Aug 14) its open-source GLM-5.3 model had neared Anthropic's restricted Mythos 5 in identifying software vulnerabilities,…

Company says GLM-5.3 outperformed leading US systems in cybersecurity tests, beating Anthropic’s Mythos 5 and OpenAI’s GPT-5.6 on CyberGym.

GLM-5.3 superou ligeiramente o Mythos 5 em um teste de segurança cibernética