GLM-5.2

Z.ai (Zhipu AI) · Open weights · Released Jun 16, 2026 · $1.40 in · $4.40 out per 1M tokens

41Overall · rank 25

Score by job

43 · rank 24 · 4 of 5 benchmarks

Benchmark	Result	Variant	z
DeepSWE	43.8%	glm-5.2_max	-0.86
FrontierCode	24.5%	glm-5.2_none	-1.56
SWE-bench Verified	78.7%	glm-5.2_max	0.70
LMArena Coding	1481 rating	glm-5.2-max	-0.45

Not enough data · 0 of 4 benchmarks

25 · rank 31 · 3 of 4 benchmarks

Benchmark	Result	Variant	z
GPQA Diamond	91.9%	glm-5.2_max	0.12
ARC-AGI-2	22.8%	glm-5.2_unknown	-1.90
SimpleBench	58.8%	glm-5.2_unknown	-0.66

20 · rank 35 · 5 of 5 benchmarks

Benchmark	Result	Variant	z
FrontierMath Tiers 1–3	59.2%	glm-5.2_max	-0.61
FrontierMath Tier 4	29.3%	glm-5.2_max	-0.49
OTIS Mock AIME	86.4%	glm-5.2_max	-2.25
ProofBench	35.0%	glm-5.2_max	-0.50
LMArena Math	1475 rating	glm-5.2-max	-0.20

64 · rank 20 · 2 of 2 benchmarks

Benchmark	Result	Variant	z
LMArena Creative Writing	1461 rating	glm-5.2-max	0.31
LMArena Instruction Following	1460 rating	glm-5.2-max	-0.19

51 · rank 16 · 1 of 1 benchmarks

Benchmark	Result	Variant	z
LMArena WebDev	1600 rating	glm-5.2-max	0.41