After a catastrophic 72 hours, who saw their most loyal users in an open uprising, Openai makes a major U-turn.
In a series of posts on X (formerly Twitter) Sunday, Director General Sam Altman announced that the company is returning its beloved older AI models, including GPT-4O, and drastically grows usage to pay subscribers, a clear peace offer to a fierce customer base.
The move comes just days after the failed gPT-5 rolling, the company’s newest and most powerful model. The launch, which should be a triumph, instead caused a fire storm.
The percentage of users using reasoning models daily is increasing significantly; For example, for free users we went from <1% to 7%, and for plus users from 7% to 24%.
I expect that using reasoning will increase greatly with time, so limit restrictions are important.
– Sam Altman (@Sama) 10 August 2025
Nasty launch and user uprising
On August 7, Openai launched GPT-5, presenting it as a “unified system”, which would automatically direct using questions to the best model for the task. In doing so, it removed the menu, which allowed users to choose from older, reliable models like GPT 4O, which was launched in March 2023.
For customers paying for subscriptions like Chatgpt Plus ($ 20/month), the sudden change felt like betrayal. They built their professional and creative flows of work around a gear of various models, each with their own strengths: one for creativity, another for pure logic, another for deep research. Forcing everyone on a single new model broke these workflows and removed their ability to pour responses to control errors or hallucinations. The reaction was immediate and fierce, resulting in a waterfall of subscription cancellations and online petitions.
Altman’s public reversal
As the outrage grew over the weekend, Sam Altman took X to make damaging control, acknowledging that the company abused the situation and announced a series of concessions.
The biggest news came in direct answer to a user asking the question in everyone’s mind: “And bringing back 4o?”
“It is coming back! Go to settings and select” shows “,” Altman replied, confirming the return of the fan-favored model. While GPT-5 remains the default, users can now choose back into the older versions.
It came back! Go to Settings and select “Show Legacy Models”
– Sam Altman (@Sama) 10 August 2025
A user then jokingly asked if they would be “marked as strange” to use the legacy models. “Not marked so strange in both cases!” Altman replied.
To further reassure paying customers, Altman has announced a massive growth of user limits for the most powerful GPT-5 features. “Today we significantly increase rhythmic limits for reasoning of Chatgpt Plus users, and all model class limits will soon be higher than they were before GPT-5,” he posted.
When a user requested specifications on the new limits, Altman revealed the new cap: “Trying 3000 a week now!” This is a huge growth for $ 20 per month -a bonus and a clear motivation for angry users to stay.
Trying 3000 a week now!
– Sam Altman (@Sama) 10 August 2025
Promise of transparency
Finally Altman promised more transparency. He announced next UI change to show users, which model actively responds to their questions and promised a detailed blog post this week explaining the “thinking about how we will make compromises on capacity.”
He also shared data showing the huge popularity of the new reasoning models, revealing that daily usage between plus users jumped from 7% to 24%, subtly justifying the company’s initial focus on the new TEchniko, even though the rolling was a disaster. “Reasoning” is the ability to “think” step by step to solve a complex problem.
In the end, Openai seems to have acquired the message loud and clear. This was a rare and powerful display of a user uprising, forcing one of the most powerful companies in TE Techniko to listen and look back. The compromise-preserving GPT-5 as the default as they give users their choice-is direct recognition of the power of a loyal community that refuses to be ignored.
