After Testing Opus 5.5 vs GPT-6 Sol, This User Is Switching Back to Claude

xeophon · x · 2026-09-23

User xeophon shares first-hand impressions after using Opus 5.5 and GPT-6 Sol on different tasks: he may switch back to Claude as his primary model for the first time since November 2025, finding its outputs more human-readable and seemingly smarter than Sol. His main gripe: Claude still tends to add too many code comments.

Original post →

More from Models

Models channel →