GLM 5 — Benchmarks

Benchmark scores for GLM 5 aggregated from public leaderboards, with how it ranks among open models. See hardware requirements for what you need to run it.

Overall rank: #18 of 123 open modelscomposite 64.5/100 across 9 benchmarks in 4 categories · methodology

Coding

BenchmarkScoreOpen rankAll models
Terminal-Bench52.4%#1 / 16#22 / 57
SWE-bench Verified72.8%#2 / 16#27 / 162
SWE-bench Multilingual69.7%#1 / 4#4 / 13

Knowledge

BenchmarkScoreOpen rankAll models
MMLU-Pro86.0%#5 / 119#30 / 259

Math

BenchmarkScoreOpen rankAll models
AIME 2024/202580.0%#22 / 74#104 / 270
FrontierMath16.4%#6 / 12#48 / 101

Reasoning

BenchmarkScoreOpen rankAll models
GPQA Diamond87.8%#14 / 83#57 / 291
ARC-AGI44.7%#10 / 16#117 / 200
SimpleBench53.2%#6 / 23#46 / 101

Scores aggregated from public benchmark sources (each linked from the benchmark pages). llmrun does not run these benchmarks.