Grok 4 vs Claude 4.5 Haiku (Reasoning) | grube.ai

Back to Benchmarks

Grok 4

xAI

vs

Claude 4.5 Haiku (Reasoning)

Anthropic

Grok 4 wins 10 of 11

Index Scores

41.5

+4.4

Intelligence

37.1

40.5

+7.9

Coding

32.6

92.7

+9.0

Math

83.7

Benchmarks

86.6

+10.6

MMLU Pro

76.0

87.7

+20.5

GPQA

67.2

23.9

+14.2

HLE

9.7

81.9

+20.4

LiveCode

61.5

45.7

+2.4

SciCode

43.3

99.0

+99.0

MATH-500

0.0

94.3

+94.3

AIME

0.0

Speed

45

TPS

+21

66

Shared

Winning delta

crafted by bart stefanski