OpenAI Revives GPT-4o After Strong User Demand
In a bold move reacting to intense user demand, OpenAI CEO Sam Altman announced the return of GPT-4o to ChatGPT users. This follows significant user backlash that greeted the removal of GPT-4o with the recent launch of GPT-5. Currently, the popular AI model is back and available by default for all paid ChatGPT users, including those with ChatGPT Plus accounts.
The decision to restore GPT-4o comes after OpenAI acknowledged they underestimated the importance of some of GPT-4o’s features to their community. With the launch of GPT-5, all previous AI models were removed in an effort to streamline their offerings, a move that inadvertently led to one of the most vocal user revolts in the platform's history. A particularly notable Reddit thread titled "GPT-5 is horrible" amassed over 2,000 comments within just days.
As part of addressing the dissatisfaction, OpenAI introduced several changes along with bringing back GPT-4o. The rate limits for GPT-5's Thinking mode have been significantly increased from 200 to 3,000 messages a week, with additional capacity through the "GPT-5 Thinking mini". Furthermore, the introduction of new routing options—"Auto," "Fast," and "Thinking"—offers users better control over the version of GPT-5 they interact with.
For premium users of ChatGPT, who invest $200 a month for enhanced access, CEO Altman did confirm future plans to make models like o3, 4.1, and GPT-5 Thinking mini available through settings. However, GPT-4.5 remains exclusively for Pro subscribers due to high operational costs.
Concerns about "personality" intrigued a broader discussion of user preferences, especially since GPT-5's output was often described as less engaging compared to GPT-4o's friendly tone. Some expressed true emotion over losing what they referred to as their "only friend" in AI. OpenAI assured users that they are striving to improve GPT-5's personality to make it warmer while maintaining user experience.
The launch of GPT-5 was not only troubled by model availability issues but also faced technical problems such as malfunctioning routing systems and misleading performance data being presented. OpenAI continues refining GPT-5, adapting rate limits based on usage trends and developing personality enhancements. Users can currently choose their preferred model, balancing OpenAI's evolution toward new AI offerings with user needs.
