Verdict
Claude Opus 5 leads 20–1 across 21 shared benchmarks.
Claude Opus 4.8 1 · Claude Opus 5 20 · higher isn't always better — check the caveats below.
Reported scores — protocols may differ. How we compare
Unknown: Claude Opus 5 leads 6–0 across 6 shared unknown benchmarks (largest gap: ProofBench, 99 vs 69 — single evaluator).
Reasoning: Claude Opus 5 leads 4–0 across 4 shared reasoning benchmarks (largest gap: SimpleBench, 76.72 vs 57.76 — single evaluator).
Agentic: Claude Opus 5 leads 2–0 across 2 shared agentic benchmarks (largest gap: APEX-Agents, 65.8 vs 48.9 — protocols undisclosed).
Coding: Claude Opus 5 leads 2–0 across 2 shared coding benchmarks (largest gap: DeepSWE, 73.65 vs 58.97 — single evaluator).
Math: Claude Opus 5 leads 2–0 across 2 shared math benchmarks (largest gap: FrontierMath-Tier-4-v2-Private, 73.17 vs 56.1 — single evaluator).
Claude Opus 5 leads 20–1 across 21 shared benchmarks. Claude Opus 4.8 leads 1 benchmark and Claude Opus 5 leads 20. Higher isn't always better — see the integrity caveats on each benchmark.
Claude Opus 4.8 and Claude Opus 5 have 21 benchmarks in common in our data — those are the only rows where a direct, apples-to-apples comparison is drawn.
Claude Opus 5 leads unknown 6–0. Claude Opus 5 leads reasoning 4–0.