OpenAI GPT-4o Returns: User Revolt Response
OpenAI Backtracks on GPT-5 launch, Reinstates Popular GPT-4o Amid user Outcry
Table of Contents
OpenAI has swiftly reversed course on its recent GPT-5 launch, bringing back the widely-loved GPT-4o model for ChatGPT Plus subscribers. This dramatic shift comes after a massive wave of complaints from users who found themselves forced onto the new model without a choice, and many disliked the changes. It’s a clear demonstration that even AI giants listen to their user base – especially when that base makes its displeasure known.
The User Revolt Against GPT-5
When OpenAI launched GPT-5 on August 7th, it did so with a sweeping change: all previous models were removed from ChatGPT. The intention was to streamline offerings and push users toward the latest and greatest. However, the execution sparked a notable backlash.
Many users found GPT-5 to be a downgrade, particularly in areas like creative writng and nuanced conversation. A Reddit thread titled “GPT-5 is horrible” quickly exploded, amassing over 2,000 comments in just a few days, filled with frustration and disappointment. Users lamented the loss of GPT-4o’s specific strengths and the lack of control over which model powered their chats.
OpenAI CEO Sam Altman acknowledged the misstep, explaining that the company had underestimated ”how much some of the things that people like in GPT-4o matter to them.” It’s a rare admission from a tech leader, and a testament to the strength of the user response.
What’s Changing: GPT-4o’s Return and Beyond
The reinstatement of GPT-4o is the most immediate change, offering users a welcome return to a familiar and preferred experience. But OpenAI didn’t stop there. Several other adjustments have been made to address the concerns raised by the community:
Increased Rate Limits for GPT-5 “Thinking” Mode: The weekly message limit for GPT-5’s more deliberate “Thinking” mode has been considerably increased from 200 to 3,000 messages. For those who hit that limit, a “GPT-5 Thinking mini“ option will provide additional capacity.
New Routing Options: OpenAI has introduced new routing options – “Auto,” “Fast,” and “Thinking” - giving you more control over which GPT-5 variant handles your requests. This allows you to tailor the AI’s approach to your specific needs, whether you prioritize speed, efficiency, or in-depth reasoning.
More Models for Pro Subscribers: For users who pay $200 a month for ChatGPT Pro access, even more models are on the way. Altman confirmed that o3, 4.1, and GPT-5 Thinking mini will become available through a “Show additional models” toggle in ChatGPT web settings.
GPT-4.5 Remains Pro Exclusive: GPT-4.5 will remain exclusive to Pro subscribers, due to the substantial GPU costs associated with running the more powerful model.
What Does This Mean for You?
This whole situation highlights the importance of user feedback in the rapidly evolving world of AI. OpenAI’s quick response demonstrates a willingness to adapt and prioritize user experience.
If you’re a ChatGPT Plus subscriber, you’ll soon have more choice and control then ever before.You can stick with the familiar GPT-4o, experiment with the different GPT-5 variants, or explore the additional models coming to Pro subscribers.