Verdict
GPT-6 Astra leads 11–2 across 14 shared benchmarks.
Claude Sonnet 5.5 2 · GPT-6 Astra 11 · 1 tied · higher isn't always better — check the caveats below.
Reported scores — protocols may differ. How we compare
Unknown: GPT-6 Astra leads 3–1 across 4 shared unknown benchmarks (largest gap: SpireClimb, 25.67 vs 1.33 — protocols undisclosed).
Reasoning: GPT-6 Astra leads 3–0 across 3 shared reasoning benchmarks (largest gap: SimpleQA Verified, 75.6 vs 46.5 — protocols undisclosed).
Coding: GPT-6 Astra leads 2–0 across 2 shared coding benchmarks (largest gap: FrontierSWE, 65.51 vs 61.89 — single evaluator).
Math: GPT-6 Astra leads 1–0 across 2 shared math benchmarks (largest gap: FrontierMath-Tier-4-v2-Private, 97.6 vs 80.49 — protocols undisclosed).
GPT-6 Astra leads 11–2 across 14 shared benchmarks. Claude Sonnet 5.5 leads 2 benchmarks and GPT-6 Astra leads 11, with 1 tied. Higher isn't always better — see the integrity caveats on each benchmark.
Claude Sonnet 5.5 and GPT-6 Astra have 14 benchmarks in common in our data — those are the only rows where a direct, apples-to-apples comparison is drawn.
GPT-6 Astra leads reasoning 3–0. GPT-6 Astra leads unknown 3–1.