GPT-5.4 (xhigh) vs Claude Opus 4.6 (Adaptive Reasoning, Max Effort) | grube.ai

Back to Benchmarks

GPT-5.4 (xhigh)

OpenAI

vs

Claude Opus 4.6 (Adaptive Reasoning, Max Effort)

Anthropic

GPT-5.4 (xhigh) wins 5 of 11

Index Scores

57.2

+4.2

Intelligence

53.0

57.3

+9.2

Coding

48.1

0.0

Math

0.0

Benchmarks

0.0

MMLU Pro

0.0

92.0

+2.4

GPQA

89.6

41.6

+4.9

HLE

36.7

0.0

LiveCode

0.0

56.6

+4.7

SciCode

51.9

0.0

MATH-500

0.0

0.0

AIME

0.0

Speed

32

TPS

+4

36

Shared

Winning delta

crafted by bart stefanski