Verdict
GPT-5.2 leads 16–5 across 21 shared benchmarks.
Claude Opus 4.5 5 · GPT-5.2 16 · higher isn't always better — check the caveats below.
Reported scores — protocols may differ. How we compare
Reasoning: GPT-5.2 leads 3–2 across 5 shared reasoning benchmarks (largest gap: SimpleBench, 54.4 vs 34.96 — single evaluator).
Agentic: Evenly split 1–1 across 2 shared agentic benchmarks (largest gap: APEX-Agents, 34.3 vs 18.4 — protocols undisclosed).
Math: GPT-5.2 leads 2–0 across 2 shared math benchmarks (largest gap: FrontierMath-Tier-4-v2-Private, 31.7 vs 4.88 — single evaluator).
GPT-5.2 leads 16–5 across 21 shared benchmarks. Claude Opus 4.5 leads 5 benchmarks and GPT-5.2 leads 16. Higher isn't always better — see the integrity caveats on each benchmark.
Claude Opus 4.5 and GPT-5.2 have 21 benchmarks in common in our data — those are the only rows where a direct, apples-to-apples comparison is drawn.
GPT-5.2 leads math 2–0. GPT-5.2 leads reasoning 3–2.