GLM-5.2 vs GPT-5.5

Verdict

GPT-5.5 leads 14–1 across 15 shared benchmarks.

GLM-5.2 1 · GPT-5.5 14 · higher isn't always better — see caveats.

Per-benchmark head-to-head

GLM-5.2GPT-5.5

35.6
APEX-Agents
38.4
77
ARC-AGI
95
22.78
ARC-AGI-2
85
21
Chess Puzzles
54
20.86
CritPt
27.14
54.6
CursorBench
64.3
29.27
FrontierMath-Tier-4-v2-Private
72.5
59.21
FrontierMath-Tiers-1-3-v2-Private
85.26
0.02
GBAEval
53.22
89.14
GPQA diamond
92
86.38
OTIS Mock AIME 2024-2025
100
34.29
PostTrainBench
25.02
38.1
SimpleQA Verified
63.1
78.7
SWE-Bench verified
80.58
70.12
WeirdML
84.91

A higher number is not always a better model. Each score is task performance under a disclosed harness, and some of these benchmarks are saturated or scored under mixed harnesses. Follow any benchmark to its integrity grade before reading a win as decisive.

Follow the record

Per-benchmark head-to-head

Follow the record

These models

GLM-5.2 vs …

GPT-5.5 vs …

Per-benchmark head-to-head

Follow the record

These models

GLM-5.2 vs …

GPT-5.5 vs …