OpenAI Brings Back Model 4.0 to ChatGPT After Backlash Against GPT-5: Users Rejoice!

The launch of GPT-5 had been highly anticipated over the recent months, yet it did not meet the expected reception. Following multiple issues, particularly with the auto-selection feature failing to choose the best model for user queries, ChatGPT users called for the reinstatement of the older models. Sam Altman, CEO of OpenAI, has listened to the user feedback and announced on Saturday, August 9, 2025, via a post on X, a phased reintroduction of the 4o model. OpenAI’s account also confirmed this announcement.

A New Model Falls Short of User Expectations

Just a day after defaulting to GPT-5 as the primary model, OpenAI reversed its decision and added GPT-4o back into the model options for ChatGPT Plus and Team members. This move came in response to widespread criticism of the new model, which users found to be slower and provided shorter, less accurate responses than GPT-4o.

We definitely underestimated how much people valued certain aspects of GPT-4o, even though GPT-5 is superior in many ways. Plus users will now have the option to continue using 4o, Sam Altman, CEO of OpenAI, wrote on X.

According to Sam Altman, the “autoswitch” feature, which allows the chatbot to select the most suitable model for a user’s request, was not functional for much of the launch day, potentially explaining the performance issues. However, he has since assured that the issues with GPT-5 have been resolved and that the model will “appear smarter from today.”

To switch back to using GPT-4o, you need to:

• Click on your profile at the bottom left of your screen,
• Press on Settings,
• In General, now check Show old models.

The redeployment of GPT-4o is already underway and may take a few days to complete.

OpenAI Increases Usage Credits for GPT-5

The CEO of OpenAI also stated that ChatGPT Pro users will receive double the credits to use GPT-5. This adjustment will allow users to explore and get accustomed to the new model, tailor it to their needs, and not worry about hitting monthly limits. Additionally, ChatGPT will enhance transparency regarding which model the “autoswitch” feature selects to respond to a query. Users will also have the option to select the model they prefer to use themselves.

We will double the usage limits for Plus users once the rollout is complete. Next week, we will deploy mini-versions of GPT-5 and GPT-5, planning to take over until your limits reset, announced Sam Altman on X.