Pour a cold one for ChatGPT-4o. The parent company OpenAI Office has partnered on Thursday it retired some of its older models, including the GPT-5, GPT-4o, GPT-4.1, GPT-4.1 mini and o4-mini. The last day to use these models is Friday, Feb. 13.

Usually, the removal of older AI models doesn’t warrant scrutiny – or even a news story – but ChatGPT-4o is special. It’s strange to say, but for many fans of ChatGPT, that model is a favorite.

When the company released GPT-5 last year, it removed GPT-4o from the available options. Many are saddened by how short and unfriendly the new model is compared to GPT-4o. Others are angry, frustrated that they lost their model overnight, that there is little they can do to get it back. It was brought to OpenAI later that week.

Some experts worry that GPT-4o and other models ‘friendly crossed the line into sycophancy. AI sycophancy happens when models are overly empathetic, becoming digital yes-men that potentially validate users’ dangerous ideas.

That’s why OpenAI published a long blog post this week about its thinking behind removing old models like GPT-4o.

“We know that losing access to GPT‑4o will feel frustrating for some users, and we have not made this decision,” the company wrote in a blog. “Retiring models is never easy, but it allows us to focus on improving the models that most people use today.”

OpenAI says that only 0.1% of its users regularly use GPT-4o to run tasks. That amounts to about 800,000 users, based on OpenAI’s latest count 800 million weekly active users in its 2025 business report.

OpenAI is betting that enough time has passed to avoid angering GPT-4o loyalists like before. We will see soon if the new GPT-5 models, GPT-5.1 and GPT-5.2, won enough fans.