User Says GPT-4o's High Reasoning Mode Still Not Smart Enough

koltregaskes · x · 2026-08-25

User koltregaskes responds to a discussion about GPT-4o's reasoning modes, saying that while a high mode was added, it's not really high and the model is still quite dumb.

Related event: Users discover GPT allows real-time reasoning adjustments(3 posts)→

Original post →

More from Models

Models channel →