Hey y’all,
I’m posting feedback about the recent removal of GPT-4o and the automatic transition to the 5.x models.
GPT-4o had a distinct interaction style that was warm, fluid, and highly responsive to conversational context. This tone meaningfully affected how many users (myself included) worked with ChatGPT for writing, planning, brainstorming, and reflective thinking. It wasn’t just a personal preference or aesthetic.
The abrupt removal of 4o, without any in-app notice, created a real disruption in daily workflows. The 5.x models are powerful, sure, but they do not replicate the same tone or conversational stability, and the shift was jarring.
I’d like to request the following:
-
Advance, in-app communication when major model changes or deprecations are planned.
-
Optional access to legacy models when they offer distinctly different interaction styles or affordances, especially for paid users.
-
Clear documentation about how safety tuning and alignment changes impact conversational tone and responsiveness.
This post is not about expecting a relationship dynamic from the model. This is about the importance of interaction style and conversational UX in a tool many people use extensively every day.
Based on conversations happening across the community, many users felt the impact of GPT-4o’s removal. I hope OpenAI will consider offering more transparency and user choice moving forward.
Thanks for reading. 