Grok 4.5

xAI · Closed weights · Released Jul 8, 2026 · $2 in · $6 out per 1M tokens

47Overall · rank 20

Score by job

53 · rank 20 · 3 of 5 benchmarks

Benchmark	Result	Variant	z
DeepSWE	53.8%	grok-4.5_high	-0.22
FrontierCode	42.4%	grok-4.5_unknown	0.30
LMArena Coding	1478 rating	grok-4.5	-0.58

Not enough data · 1 of 4 benchmarks

Benchmark	Result	Variant	z
APEX-Agents	56.2%	grok-4.5_unknown	0.37

56 · rank 15 · 3 of 4 benchmarks

Benchmark	Result	Variant	z
GPQA Diamond	93.4%	grok-4.5_high	0.76
ARC-AGI-2	52.6%	grok-4.5_high	-0.52
SimpleBench	70.0%	grok-4.5_unknown	0.46

31 · rank 29 · 5 of 5 benchmarks

Benchmark	Result	Variant	z
FrontierMath Tiers 1–3	57.2%	grok-4.5_high	-0.74
FrontierMath Tier 4	24.4%	grok-4.5_high	-0.69
OTIS Mock AIME	97.8%	grok-4.5_high	0.54
ProofBench	31.0%	grok-4.5_high	-0.66
LMArena Math	1462 rating	grok-4.5	-0.69

54 · rank 30 · 2 of 2 benchmarks

Benchmark	Result	Variant	z
LMArena Creative Writing	1440 rating	grok-4.5	-0.32
LMArena Instruction Following	1449 rating	grok-4.5	-0.56

40 · rank 22 · 1 of 1 benchmarks

Benchmark	Result	Variant	z
LMArena WebDev	1552 rating	grok-4.5	-0.05