7.2
/ 10
GLM-5.2
98 USER VOICES · REALITY SCORE
Open-source LLM from Z.ai positioning itself near frontier, but real-world tests show reasoning-efficiency and cost-per-intelligence gaps vs