Claude Opus 5.5 hits #2 on FrontierSWE at 62.3%, trailing GPT-6 Astra's 65.5%

scaling01 · x · 2026-09-23

Per the FrontierSWE leaderboard from ProximalHQ, Claude Opus 5.5 scores 62.3%, ranking #2 — just behind GPT-6 Astra (65.5%) and clearly ahead of Fable 5.1 (56.3%) and the previous Opus 5 (52.0%).

Original post →

More from Models

Models channel →