Claude Opus 5.5 vs GPT-6 Sol

Verdict

Claude Opus 5.5 leads 14–2 across 17 shared benchmarks.

Claude Opus 5.5 14 · GPT-6 Sol 2 · 1 tied · higher isn't always better — check the caveats below.

Per-benchmark head-to-head

Reported scores — protocols may differ. How we compare

Sort

Claude Opus 5.5GPT-6 Sol

73.5
AAPEX-Agents
54.3
98.5
BARC-AGI
95.5
93.33
AARC-AGI-2
89.58
65.28
AChess Puzzles
53.7
98.22
ADTBench
95.55
71.43
AEBR-bench
53.33
54.64
AFrontierCode
49.27
95
AFrontierMath-Tier-4-v2-Private
90
91.23
AFrontierMath-Tiers-1-3-v2-Private
89.82
76.19
AFurniture Assembly
40.48
87.46
BGPQA diamond
92.42
80.27
ALMCA
69.51
68.05
AMystery Game Puzzles
51.53
100
BOTIS Mock AIME 2024-2025
100
100
BProofBench
83
72.2
ASimpleQA Verified
60.7
3.33
ASpireClimb
7.5

Capabilities

Unknown

Unknown: Claude Opus 5.5 leads 6–1 across 7 shared unknown benchmarks (largest gap: Furniture Assembly, 76.19 vs 40.48 — protocols undisclosed).

14 unverified

Reasoning

Reasoning: Claude Opus 5.5 leads 2–1 across 3 shared reasoning benchmarks (largest gap: SimpleQA Verified, 72.2 vs 60.7 — protocols undisclosed).

6 unverified

Math

Math: Claude Opus 5.5 leads 1–0 across 2 shared math benchmarks (largest gap: FrontierMath-Tier-4-v2-Private, 95 vs 90 — protocols undisclosed).

4 unverified

A higher score does not always mean a better model. Each number is task performance under a disclosed harness. Of the 34 measurements across 17 shared benchmarks, 34 single-evaluator or undisclosed-protocol reports; 0% are independently reproduced. Check any benchmark's integrity grade before reading a win as decisive.

Frequently asked questions

Claude Opus 5.5 leads 14–2 across 17 shared benchmarks. Claude Opus 5.5 leads 14 benchmarks and GPT-6 Sol leads 2, with 1 tied. Higher isn't always better — see the integrity caveats on each benchmark.

Claude Opus 5.5 and GPT-6 Sol have 17 benchmarks in common in our data — those are the only rows where a direct, apples-to-apples comparison is drawn.

Claude Opus 5.5 leads unknown 6–1. Claude Opus 5.5 leads reasoning 2–1.

Follow the record

These models

Same tier