Claude Opus 4.8 vs Gemini 3.1 Pro — Who Actually Wins? Real Benchmark Data (2026)

head to head

Claude Opus 4.8 vs Gemini 3.1 Pro

13 shared benchmarks

7

wins

0

ties

6

wins

95.7%

AIME

98.13%

1.5%

ARC-AGI 3

0.37%

84.3%

BrowseComp

85.9%

80.5%

CharXiv

83.3%

59.0%

DeepSWE

12.0%

80.0%

FrontierMath

36.9%

91.0%

GPQA Diamond

94.3%

49.8%

Humanity's Last Exam

44.4%

77.8%

MCP-Atlas

78.2%

78.9%

MMMU Pro

80.5%

84.0%

Online-Mind2Web

57.3%

69.2%

SWE-bench Pro

54.2%

84.6%

Terminal-Bench 2.1

70.3%