OpenAI to restore older ChatGPT models amid GPT-5 criticism

ChatGPT users are complaining that GPT-5 responses are less thorough, shorter, and lack originality
An undated image. — OpenAI
An undated image. — OpenAI

In a major development, OpenAI is considering reintroducing the GPT-4o model for ChatGPT Plus subscribers following widespread dissatisfaction with its latest GPT-5 release, CEO Sam Altman confirmed.

Described as OpenAI’s most advanced and capable model to date, the launch of GPT-5 has been met with mixed reviews.

Many users have taken to social media, particularly Reddit, to express frustration over what they see as less detailed, shorter, and less engaging responses compared to GPT-4o.

Some described GPT-5 as “cold” and “bland”, with one viral comment urging, “BRING BACK 4o.”

Another compared the experience to “wearing the skin of my dead friend”, highlighting the stark difference in tone and output.

Altman acknowledged the criticism, stating that the company is exploring options to let Plus subscribers continue using GPT-4o.

However, he noted that supporting multiple high-tier models comes with technical and operational costs.

“We hear you all on 4o,” he said, adding that usage patterns will determine how long the model remains available.

He also acknowledged the GPT-5 rollout was “a little more bumpy than we hoped,” citing a temporary malfunction in the system that determines whether prompts are handled by a more intensive reasoning model or a lighter, faster one.