Verdict
GPT-5.6 Sol leads 20–1 across 21 shared benchmarks.
Claude Opus 4.8 1 · GPT-5.6 Sol 20 · higher isn't always better — check the caveats below.
Reported scores — protocols may differ. How we compare
Unknown: GPT-5.6 Sol leads 6–0 across 6 shared unknown benchmarks (largest gap: Mystery Game Puzzles, 53.73 vs 29.5 — protocols undisclosed).
Reasoning: GPT-5.6 Sol leads 4–0 across 4 shared reasoning benchmarks (largest gap: SimpleQA Verified, 69.7 vs 53 — protocols undisclosed).
Agentic: Evenly split 1–1 across 2 shared agentic benchmarks (largest gap: OSWorld 2.0, 27.34 vs 20.6 — single evaluator).
Coding: GPT-5.6 Sol leads 2–0 across 2 shared coding benchmarks (largest gap: DeepSWE, 72.67 vs 58.97 — single evaluator).
Math: GPT-5.6 Sol leads 2–0 across 2 shared math benchmarks (largest gap: FrontierMath-Tier-4-v2-Private, 82.93 vs 56.1 — single evaluator).
GPT-5.6 Sol leads 20–1 across 21 shared benchmarks. Claude Opus 4.8 leads 1 benchmark and GPT-5.6 Sol leads 20. Higher isn't always better — see the integrity caveats on each benchmark.
Claude Opus 4.8 and GPT-5.6 Sol have 21 benchmarks in common in our data — those are the only rows where a direct, apples-to-apples comparison is drawn.
GPT-5.6 Sol leads unknown 6–0. GPT-5.6 Sol leads reasoning 4–0.