January 29, 2026 – On Thursday, OpenAI made an announcement that it would be discontinuing the use of several models in its ChatGPT chatbot starting next month, among which the well – liked GPT – 4o model by many users is included.
Back in May 2024, OpenAI launched the GPT – 4o large – scale model. Its friendly conversational style quickly won the favor of a portion of paying users. However, in August of the same year, after introducing the new model GPT – 5, OpenAI briefly shut down access to GPT – 4o, which sparked a wave of dissatisfaction among users.
In response to the backlash, OpenAI promptly restored access to GPT – 4o for paying users. CEO Sam Altman promised that if the company ever decided to discontinue the model in the future, it would give users sufficient prior notice.
Well, that day has arrived. On Thursday, OpenAI stated that currently, only 0.1% of daily users choose to use GPT – 4o, while the “vast majority” are using its GPT – 5.2 model. OpenAI explained that over the past few months, the company has made improvements in model personality, customization, and creative ideation. Based on these developments, it believes the time is right to officially phase out GPT – 4o.
In addition to GPT – 4o, OpenAI also announced that GPT – 4.1, GPT – 4.1 mini, and OpenAI o4 – mini would also be removed from ChatGPT. Previously, the company had already declared that GPT – 5 Instant and GPT – 5 Thinking would be taken out of the chatbot as well.
