GPT-5.2 vs GPT-5.3 Codex — Who Actually Wins? Real Benchmark Data (2026)

head to head

GPT-5.2 vs GPT-5.3 Codex

3 shared benchmarks

1

wins

1

ties

1

wins

92.4%

GPQA Diamond

91.5%

27.8%

Humanity's Last Exam

39.9%

80.0%

SWE-bench Verified

80.0%