After facing intense backlash from its most dedicated users over the past 72 hours, OpenAI has made a significant reversal. CEO Sam Altman announced on X (formerly Twitter) that the company will restore popular older AI models, including GPT-4o, and will significantly raise usage limits for paying subscribers.
This decision aims to address the frustrations of a frustrated customer base following the problematic launch of GPT-5, OpenAI’s newest and most advanced model. What was expected to be a smooth rollout quickly turned into widespread dissatisfaction, prompting OpenAI to act swiftly to regain user trust and satisfaction.
Read More: Top 6 Low-Sodium Foods for Better Heart Health
A Failed Launch Sparks User Revolt
On August 7, OpenAI introduced GPT-5 as a “unified system” designed to automatically direct queries to the most suitable model. This update removed the option for users to select trusted older models like GPT-4o, released in March 2023.
Subscribers to services like ChatGPT Plus ($20/month) felt betrayed by this sudden change. Many had relied on multiple models for different tasks—creativity, logic, and research—and the forced switch disrupted their workflows. Without the ability to cross-check answers across models, users lost a critical tool for verifying accuracy. The backlash was swift and intense, triggering a wave of subscription cancellations and online petitions demanding the return of the previous system.
Altman’s Public Reversal
As user outrage escalated over the weekend, Sam Altman addressed the backlash on X, admitting the company misjudged the rollout and announcing key concessions.
In response to a user’s question about GPT-4o, Altman confirmed its return: “It’s back! Go to settings and pick ‘show legacy models.’” While GPT-5 remains the default, users can now choose older models again. When asked if using legacy models would be “marked as weird,” Altman assured, “Not marked as weird in either case!”
To further satisfy subscribers, Altman revealed a major boost in usage limits for GPT-5’s advanced features. He shared, “We are significantly increasing rate limits for reasoning for ChatGPT Plus users, with all model-class limits soon higher than before GPT-5.” Specifically, users can now enjoy up to 3,000 requests per week—a substantial upgrade for a $20/month plan, encouraging frustrated users to stay.
A Commitment to Transparency
Altman also pledged greater transparency, announcing an upcoming interface update that will show users which model is responding to their queries. He promised a detailed blog post this week outlining the company’s approach to managing capacity trade-offs.
Sharing usage data, Altman revealed that daily engagement with the new reasoning models among Plus users surged from 7% to 24%, highlighting the appeal of the advanced technology despite the rocky rollout. “Reasoning” refers to the AI’s step-by-step problem-solving ability.
Ultimately, OpenAI responded decisively to user demands. By keeping GPT-5 as the default while restoring user choice, the company acknowledged the influence of its loyal community, proving that even tech giants must listen when their users speak up.
Frequently Asked Questions
Why did OpenAI bring back GPT-4o?
OpenAI restored GPT-4o after strong user backlash over the GPT-5 rollout, allowing users to choose their preferred AI models again and maintain their established workflows.
What changes were made to GPT-5 usage limits?
OpenAI significantly increased usage limits for GPT-5’s advanced features, now allowing ChatGPT Plus users up to 3,000 requests per week, improving accessibility for paying subscribers.
How can I switch between GPT-5 and older models like GPT-4o?
Users can enable legacy models by going to settings and selecting “show legacy models,” giving them the option to choose which AI model to use.
Will using legacy models affect my experience or status?
No, users won’t be marked differently or limited when using older models; OpenAI encourages choice without stigma.
What steps is OpenAI taking to improve transparency?
l is responding and will release a detailed blog explaining capacity management decisions.
Conclusion
OpenAI’s swift response to user feedback highlights the company’s commitment to balancing innovation with customer needs. By reinstating GPT-4o and boosting GPT-5 usage limits, OpenAI restores user choice and trust while continuing to push AI capabilities forward. This episode underscores the power of a loyal community and the importance of transparency in technology development.
