§ models · storyline
Z.ai GLM-5.3 beats Mythos 5 on CyberGym, restricts access
Z.ai releases GLM-5.3 model scoring 84.5% on CyberGym versus Mythos 5's 83.8%, with cybersecurity features restricted to verified users.
Reuters: Z.ai says GLM-5.3 scores 84.5% on CyberGym, vs. Mythos 5's 83.8%, and its most sensitive cybersecurity functions will only be available to verified users — Chinese AI startup Z.ai said on Friday its open-source GLM-5.3 model had neared Anthropic's restricted Mythos 5 in identifying software …
§ sources3 publications · timeline below
- reuters.comZ.ai says GLM-5.3 scores 84.5% on CyberGym, vs. Mythos 5's 83.8%, and its most sensitive cybersecurity functions will only be available to verified users (Reuters)primary
- interconnects.aiGLM-5.3: How Chinese labs keep stride with the frontier
- marktechpost.comZ.ai Ships GLM-5.3 Without Retraining the Base Model
§ how this story moved
- primary — MarkTechPost publishes the launch post.
- Interconnects (Nathan Lambert) picks up coverage.
- Reuters picks up coverage.