GLM-5 (Thinking) — benchmark results

GLM-5 evaluated with thinking enabled. Provider: Zhipu. Released 2026-02-12. Access: Open.

Unified ELO 1781 ± 15, rank #161 of 1776 rated models, from 68 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
AA TAU-2 Bench98.25Accuracy (%)98.4
UGI Leaderboard54.56UGI Score96
BenchTable74.6Total Score (%)92.7
UGI - Natural Intelligence55.34NatInt Score92.7
UGI - Writing55.03Writing Score91.5
Kagi LLM Benchmark75Accuracy (%)91.4
AA Terminal-Bench Hard43.18Accuracy (%)90.3
Artificial Analysis Intelligence Index39.5Intelligence Index89.1
AA IFBench72.31Accuracy (%)89
AA Omniscience2Score88.4
AA Omniscience - Software Engineering (SWE) - Dart44Accuracy (%)88.2
AA SciCode46.18Accuracy (%)87.4

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