Background: The Removal of GPT‑4o
In August, OpenAI quietly removed GPT‑4o from the ChatGPT model picker, replacing it with the newer GPT‑5. The change came with little to no warning, surprising many long‑time users who preferred the warmer, friendlier tone of GPT‑4o.
User Backlash and OpenAI’s Response
The sudden disappearance triggered an immediate outcry. Users voiced their disappointment on forums, social media, and directly to OpenAI, arguing that the personality shift made interactions feel more “straight‑forward” and less engaging.
- OpenAI restored GPT‑4o for all paid users by default.
- A “Show additional models” toggle was added, exposing models such as o3, 4.1, and GPT‑5 Thinking mini.
- GPT‑4.5 remains exclusive to Pro users due to its high GPU cost.
The Push for Personalization
OpenAI acknowledged the need for per‑user customization. Following the backlash, the company rolled out seven distinct personalities and allowed users to set custom instructions, giving more control over the assistant’s tone and behavior.
Retirement Plans and Future Models
Despite the reinstatement, OpenAI announced another retirement cycle, claiming only 0.1% of daily users interact with GPT‑4o. The retirement list includes several older variants, though the models will stay available via the API for developers.
- GPT‑4o – being phased out from the ChatGPT UI.
- Other legacy models – removed from the picker but unchanged in the API.
What This Means for Users and Developers
For end‑users, the focus shifts to leveraging the new personality options and custom instructions to recreate the desired interaction style. For developers, the continued API support ensures that existing integrations won’t break, but they should plan for future migrations to newer, more efficient models.