Beginning this Friday, OpenAI will discontinue access to five older ChatGPT models, including the widely used yet contentious GPT-4o. This particular model has been involved in multiple legal cases related to user self-harm, delusional conduct, and AI psychosis. It continues to hold OpenAI’s top rating for sycophancy.

Alongside GPT-4o, the GPT-5, GPT-4.1, GPT-4.1 mini, and OpenAI o4-mini models are also being phased out. OpenAI originally planned to retire GPT-4o in August when it introduced the GPT-5 model. However, significant pushback at that time led the company to maintain the legacy model for paid subscribers, who retained the option to manually select it for interactions.

In a recent blog post, OpenAI highlighted that only 0.1% of customers have been utilizing GPT-4o. For a company with 800 million weekly active users, however, that modest percentage still represents approximately 800,000 individuals. Thousands of users have voiced opposition to discontinuing the 4o model, pointing to their strong personal connections with it.