GLM-4.5 (Thinking) — benchmark results

GLM-4.5 evaluated with thinking enabled. Provider: Zhipu. Released 2025-07-28. Access: Open.

Unified ELO 1658 ± 23, rank #352 of 1776 rated models, from 13 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
LLM2014 Code 2025-09 - TypeScript8.89Score92.1
LLM2014 Code 2025-09 - C++5.62Score84.2
LLM2014 Code 2025-09 - Python7.15Score78.9
BenchTable60.8Total Score (%)71.7
LLM2014 Logic 2025-0842.18Median Score59.1
LLM2014 Code 2025-09 - C#6.94Score57.9
LLM2014 Code 2025-0937.21Median Score55.6
LisanBench0.03Mean Path Length / Current Maximum55
LLM2014 Logic 2025-0937.61Median Score51.1
Epoch AI - Algotune1.52Score50
WeirdML40.58Average Score40.1
LLM2014 Code 2025-09 - Golang3.79Score31.6

Interactive version: theaggregate.ai/model?slug=glm-4-5-thinking · How the rankings work · Data refreshed daily, snapshot 2026-07-22.