Benchmark chart pits Opus 5.5 against GPT-6 Sol

Itsvishal @itsvishaltwtOriginally posted 2026.09.23

Open original on X

Claude Opus 5.5 vs GPT-6 Sol Claude wins almost every benchmark — smarter, better at coding & automation. Claude wins most benchmarks: • Intelligence: 58 vs 48 • Coding: 54.4% vs 49.3% • Automation: 40% vs 33.2% But GPT-6 uses way fewer tokens and costs half as much.