GLM-5.3

Z.ai (Zhipu AI) · Closed weights · Released Aug 14, 2026 · $1.40 in · $4.40 out per 1M tokens

60Overall · rank 10

Score by job

70 · rank 11 · 3 of 5 benchmarks

Benchmark	Result	Variant	z
DeepSWE	69.0%	glm-5.3_max	0.76
FrontierCode	40.1%	glm-5.3_max	0.06
LMArena Coding	1500 rating	glm-5.3-max	0.46

Not enough data · 1 of 4 benchmarks

Benchmark	Result	Variant	z
APEX-Agents	56.6%	glm-5.3_unknown	0.41

Not enough data · 1 of 4 benchmarks

Benchmark	Result	Variant	z
GPQA Diamond	90.9%	glm-5.3_max	-0.26

40 · rank 21 · 5 of 5 benchmarks

Benchmark	Result	Variant	z
FrontierMath Tiers 1–3	68.8%	glm-5.3_max	0.05
FrontierMath Tier 4	29.3%	glm-5.3_max	-0.49
OTIS Mock AIME	91.1%	glm-5.3_max	-1.09
ProofBench	49.0%	glm-5.3_max	0.05
LMArena Math	1498 rating	glm-5.3-max	0.68

72 · rank 9 · 2 of 2 benchmarks

Benchmark	Result	Variant	z
LMArena Creative Writing	1462 rating	glm-5.3-max	0.34
LMArena Instruction Following	1481 rating	glm-5.3-max	0.61

56 · rank 11 · 1 of 1 benchmarks

Benchmark	Result	Variant	z
LMArena WebDev	1622 rating	glm-5.3-max	0.63