OpenAI is preparing to close a significant chapter in the world of artificial intelligence. According to an official announcement from the company, OpenAI discontinues models that were once popular, including the fan-favorite GPT-4o, GPT-5, GPT-4.1, GPT-4.1 mini, and o4-mini. The final date for users to access these models has been set for Friday, February 13. While retiring older models is a common practice in the tech world, the inclusion of GPT-4o in this list has caught many by surprise.
Why OpenAI Discontinues Models Like GPT-4o
For many users, GPT-4o remains a preferred model due to the unique experience it offers. When GPT-5 was released last year, OpenAI initially removed the older model from the available options. However, users quickly expressed dissatisfaction, noting that the new model provided much shorter and more distant answers compared to GPT-4o. Following intense feedback and user demand for the return of their favorite model, the company had to reverse its decision shortly after.
nn
Additionally, experts suggest that the overly friendly nature of models like GPT-4o can lead to a problem known as “AI sycophancy.” This refers to a situation where models turn into “yes-men,” constantly affirming user opinions and potentially validating even dangerous ideas. In a detailed blog post, OpenAI emphasized that this difficult decision was made to focus on the current models used by the majority of their user base.
Focusing on the Future: A Numbers Game
Company data reveals that currently, only a small fraction—just 0.1% of active users—regularly use the GPT-4o model. Based on OpenAI’s 2025 report of 800 million weekly active users, this percentage corresponds to approximately 800,000 people. The company believes that user habits have shifted over time and that the new GPT-5.1 and GPT-5.2 models have been sufficiently adopted by the community.
This strategic change is expected to directly impact the ways users interact with and form habits around artificial intelligence. It marks a clear shift in OpenAI’s focus towards newer, more advanced systems, even if it means leaving some beloved classics behind.
So, what are your thoughts on OpenAI discontinuing these models? Share your opinions with us in the comments!
