// freeCodeCamp · 13 April 2026

GPT-5.4 vs GLM-5: Is Open Source Finally Matching Proprietary AI?

On March 27, 2026, Zhipu AI quietly pushed an update to their open-weight model line. GLM-5.1, they claim, now performs at 94.6% of Claude Opus 4.6 on coding benchmarks. That's a 28% improvement over

freeCodeCamp

@freecodecamp · Oyedele Tioluwani

www.freecodecamp.org

freeCodeCamp@freecodecamp