GLM-4.5 — benchmark results

Zhipu's open GLM-4.5 MoE model for reasoning, coding, and agentic tasks (355B total, 32B active). Provider: Zhipu. Released 2025-07-28. Access: Open.

Unified ELO 1642 ± 14, rank #389 of 1776 rated models, from 112 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
YapBench1427YapIndex (lower is better)100
ZeroEval MATH-50098.2MATH-500 Score93.5
AI for Education Pedagogy - Social studies87.27Accuracy (%)93.2
SEAL - Fortress59.58Score90.9
AI for Education Pedagogy - Technology85.85Accuracy (%)90.5
LLM Stats (AIME 2024)91Score (%)84.9
WebCoderBench - Visual Experience87.42Score (%)84.6
AI for Education SEND81.19Accuracy (%)84.2
LiveMedBench22.46Overall Score (%)83.8
AI for Education Pedagogy - Primary90.61Accuracy (%)82
AI for Education Pedagogy - Science90.16Accuracy (%)81.8
MathArena - AIME 202593.33Accuracy (%)80.6

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