OpenAI Lags Behind Claude in Writing Experience

RaphaelDabadie · x · 2026-07-13

The author believes Sol and Fable represent a new round of model capability improvements, but the instruction-following experience in 5.6 is disappointing for writing tasks. Meanwhile, they still feel Claude outperforms OpenAI in coherence and clarity, though both models tend to overuse unnecessary jargon. The author concludes by urging OpenAI: bring back GPT-4.5.

Original post →

More from Models

Models channel →