GLM-5.2 — benchmark results

Zhipu's flagship GLM-5.2 model for reasoning, coding, and agentic tasks. Provider: Zhipu. Released 2026-06-13. Access: Open.

Unified ELO 1837 ± 12, rank #109 of 1776 rated models, from 144 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
Guesswork0.66MAE (z-score units)100
LLM Stats (AIME 2026)99.2Score (%)100
LLM Stats (NL2Repo)48.9Score (%)100
Tau3-Bench Airline87.5Pass@1 (%)100
Tau3-Bench Retail85.7Pass@1 (%)100
Tau3-Bench Telecom99.3Pass@1 (%)100
Design Arena (Website)1338Elo98.7
Design Arena (Game Dev)1350Elo98.6
OpenClawProBench81.3Overall Score (%)98.5
Design Arena (3D)1352Elo97.7
CritPt16.7Accuracy (self-reported)96.9
Design Arena (UI Components)1335Elo96.3

Interactive version: theaggregate.ai/model?slug=glm-5-2 · How the rankings work · Data refreshed daily, snapshot 2026-07-22.