OpenAI Retires GPT-4.5 and o3 from ChatGPT Despite Nostalgia

OpenAI Announces the End of GPT-4.5 and o3 on ChatGPT

Two popular artificial intelligence models will soon be removed from the ChatGPT platform. According to the announcement made by OpenAI, the reasoning model o3 will be withdrawn from the application on August 26, following a 90-day transition period. Additionally, GPT-4.5 will be removed from ChatGPT on June 27, after a 30-day transition period. These changes apply only to ChatGPT and do not affect the API.

The announcement is not surprising, as OpenAI has already launched newer, more efficient models for its users. In the free version, GPT-4.5 has long been replaced. In the paid version, users must utilize a model selector to access this older AI or the o3 model, specifically dedicated to tasks requiring advanced reasoning. However, soon, GPT-4.5 and o3 will be inaccessible, even for ChatGPT Plus subscribers.

Justification and User Reactions

OpenAI's decision is understandable, given that it already offers AIs that are theoretically more efficient than GPT-4.5 and o3. "We continue to retire older, less-used models in ChatGPT to better serve our newer, more efficient models," OpenAI justifies in its announcement.

However, despite the availability of more advanced technologies, some users remain attached to the older versions, particularly GPT-4.5. A less powerful AI, but still useful. On X, a user named Striver states, "To this day, 4.5 is the best writing model. o3 was a native pure reasoning model. 5 series still doesn’t match what those two had."

Persistent Nostalgia

Tibor Blaho, who closely follows the evolution of AI chatbots, also regrets the end of GPT-4.5, which was his favorite AI model. However, he has stopped using it because he found it too slow: "OpenAI is retiring o3 from ChatGPT on August 26, 2026 and GPT-4.5 on June 27, 2026 (these changes apply only to ChatGPT, not the API). I will miss GPT-4.5, as it was my favorite model, but I have to admit that I stopped using it a while ago because it was too slow."

Luis Mejia, another user, is trying to recreate the writing style of GPT-4.5 using newer AIs, expressing nostalgia for the unique features of these older models: "I’m trying to recreate GPT-4.5 style of writing with the power of 5.5 pro a man can dream."

A Precedent at OpenAI

All of this recalls the controversy that OpenAI had to manage when it announced the end of its GPT-4o model in 2025. In response to the dissatisfaction of many users, the company even backtracked, allowing ChatGPT Plus users to revert to GPT-4o despite the availability of a newer AI. OpenAI also had to make adjustments to GPT-5, the replacement for GPT-4o, as users found the original version of this new AI too cold.

In conclusion, while OpenAI's decision is motivated by the continuous improvement of its technologies, it elicits mixed reactions among users, some lamenting the disappearance of models to which they were attached.