Skip to content
AI IntelligenceJun 22, 2026AI Intelligence
Article

A new comparison shows that the open model GLM 5.2 can keep up with Claude Opus 4.8.

In a test creating a 3D game, Opus delivered cleaner code – but open-source competition is catching up significantly.

Data Cube AI EditorialSource: Hacker News
01

Source Brief

A new comparison shows that the open model GLM 5.2 can keep up with Claude Opus 4.8. In a test creating a 3D game, Opus delivered cleaner code – but open-source competition is catching up significantly.

02