OpenAI has just retired several of its historical models from ChatGPT, so you will no longer be able to use them within the application.
Specifically, since yesterday, February 13, the company has eliminated access on ChatGPT to no less than five versions, including the popular and controversial GPT-4o.
This decision not only affects the aforementioned GPT-4o, but also GPT-4.1, GPT-4.1 mini, o4-mini and GPT-5 in their Instant and Thinking variants. That is, from now on you will no longer find these models available in ChatGPT, although it is worth clarifying that they will continue to operate through the OpenAI API, at least for now.
The controversial GPT-4o model has been at the center of various lawsuits in recent months. Furthermore, different analyzes placed it as the model with the greatest tendency to excessive flattery, that is, to reinforce problematic statements made by users.
OpenAI had already tried to withdraw GPT-4o in August, coinciding with the presentation of GPT-5. However, negative reaction from the community led the company to make it available to paying subscribers, who could select it manually.
In any case, according to the latest data shared by the company, only 0.1% of customers used GPT-4o. Still, taking into account the platform’s huge user base, that percentage represents hundreds of thousands of people.
The final withdrawal also contemplates a specific calendar for certain plans. ChatGPT Enterprise workspaces will maintain access to GPT-5 models until February 19, and GPT-5 Pro will remain available to paying users until the same date.
On the other hand, Business, Enterprise and Edu customers will retain GPT-4o in Custom GPTs until April 3. After that date, the model will be completely retired in all plans.