OPENAI brings back the favorite GPT-4O of fans after a massive user revolt

After 72 disastrous hours that saw its most loyal users in open revolt, Openai takes a major tour.

In a series of messages on X (formerly Twitter) on Sunday, CEO Sam Altman announced that the company brought back its older beloved AI models, including GPT-4O, and considerably increases the limits of use for the payment of subscribers, a clear peace offer to furious customers.

This decision comes only a few days after the sloppy deployment of GPT-5, the most recent and powerful model of the company. The launch, which should have been a triumph, rather sparked a storm of fire.

The percentage of users using models of reasoning each day increases considerably; For example, for free users, we went from <1% to 7% and for users from more than 7% to 24%.

I expect the use of reasoning to increase considerably over time, so that the rate limit increases are significant.

– Sam Altman (@sama) August 10, 2025

A sloppy launch and a user mutiny

On August 7, Openai launched GPT-5, presenting it as a “unified system” which would automatically transport user requests to the best model for work. In doing so, he deleted the menu that allowed users to choose between older models of trust like GPT-4O, which was launched in March 2023.

For customers who pay subscriptions like Chatgpt Plus ($ 20 / month), sudden change looked like betrayal. They had built their professional and creative workflows around a toolbox of different models, each with its own forces: one for creativity, another for pure logic, another for deep research. Forcing everyone on a single new model has broken these workflows and has deleted their ability to reference the answers to check errors or hallucinations. The backlash was immediate and ferocious, which led to a cascade of subscription and petitions cancellations online.

Altman public reversal

As indignation increased during the weekend, Sam Altman went to X to do the control of the damage, recognizing that the company had badly judged the situation and announced a series of concessions.

The biggest news came in a direct answer to a user asking him: “And bring back 4o?”

“It’s back!” Go to Settings and choose “Show Legacy Models”, “replied Altman, confirming the return of the favorite fans. While GPT-5 remains the default, users can now resume older versions.

It’s back! Access the settings and choose “Show Legacy Models”

– Sam Altman (@sama) August 10, 2025

A user then asked if he would be “marked as strange” for the use of inherited models. “Not marked as weird in both cases!” Altman answered.

To further appease paid customers, Altman has announced a massive increase in the limits of use of GPT-5’s most powerful features. “Today, we considerably increase the rate limits for reasoning for Chatgpt Plus users, and all model class limits will soon be higher than before GPT-5,” he published.

When a user has asked for details on the new limits, Altman revealed the new ceiling: “Try 3000 per week now!” This is a huge increase for a subscription of $ 20 per month and a clear incentive for angry users to stay.

Try 3000 per week now!

– Sam Altman (@sama) August 10, 2025

A promise of transparency

Finally, Altman has promised more transparency. He announced an upcoming user interface change to show users which model actively responds to their requests and promised a detailed blog article this week explaining “company reflection on how we are going to make capacity compromises”.

He also shared data showing the immense popularity of new reasoning models, revealing that daily use among users has increased from 7% to 24%, subtly justifying the company’s initial emphasis on new technology, even if deployment was a disaster. The “reasoning” is the capacity of the AI to “think” step by step to solve a complex problem.

In the end, Openai seems to have received the strong and strong message. It was a rare and powerful display of user revolt, forcing one of the most powerful companies in the technology to listen to and go back. The compromise – by keeping GPT -5 as by default while giving users their choice – is a direct recognition of the power of a faithful community which refuses to be ignored.