GLM-4.5 (Non-reasoning): benchmark results

Provider: Zhipu. Released 2025-07-28. Access: Open.

Unified ELO 1608 ± 1, rank #580 of 3078 rated models, from 22 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
UGI Leaderboard58.64UGI Score98.3
ReLE - Agents - BFCL-V370.8Accuracy (%)90.3
UGI - Writing54.18Writing Score90.1
ReLE - Agents and Tool Use67.9Accuracy (%)88.7
UGI - Natural Intelligence38.72NatInt Score85.8
ReLE - Education - Primary School Subjects60.7Accuracy (%)61.5
ReLE - Reasoning - BBH79.5Accuracy (%)55.1
ReLE - Education - Gaokao46.4Accuracy (%)48.3
ReLE - Overall61.8Accuracy (%)46
ReLE - Medicine - Physician Exams78.5Accuracy (%)42.1
ReLE - Education45.2Accuracy (%)40.4
ReLE - Language and Instruction Following61.6Accuracy (%)39.9

Interactive version: theaggregate.ai/model?slug=glm-4-5-non-reasoning · How It Works · Data refreshed daily, snapshot 2026-09-19.