OpenAI has announced major updates to ChatGPT, including the return of GPT-4o as a legacy model and the introduction of new performance modes for GPT-5. These changes respond directly to user feedback, calling for greater flexibility in model selection and settings that adapt to speed and depth requirements.

After being removed from the main model list, GPT-4o is now back by default in the “Legacy models” category for Plus and Team subscribers. This decision came after many users expressed strong emotional attachment to GPT-4o, often describing it as a conversation partner with a unique style.

Users can now access GPT-4o directly without additional settings, or enable the “Show additional models” option to add others like o3, GPT-4.1, and GPT-5 Thinking mini to their model picker.

Another key update is the introduction of three main modes in GPT-5: Auto, Fast, and Thinking.

• Auto automatically chooses between speed and depth based on the query.

• Fast prioritizes quick responses with minimal processing time.

• Thinking takes longer but provides more detailed and structured answers.

In addition, there is Thinking mini, a lighter and faster version of Thinking, available when the weekly Thinking limit is reached.

OpenAI has also raised the weekly limit for Plus and Team users on GPT-5 Thinking mode to up to 3,000 messages per week. Once this cap is reached, users can continue their work with GPT-5 Thinking mini without waiting for the weekly reset. This change is expected to give greater flexibility, especially to users with intensive needs such as research and content development.

Starting this week, GPT-5 is available for Enterprise and Edu customers, extending access to large corporations and educational institutions. This expansion is aimed at driving wider AI adoption in professional and academic environments, strengthening OpenAI’s market position in strategic sectors.

OpenAI has confirmed plans to release a warmer and more familiar personality version of GPT-5. This tonal shift is designed to make interactions feel more natural, addressing feedback that the initial GPT-5 felt more formal and “cold” compared to GPT-4o.

The move reflects OpenAI’s commitment to enhancing the user experience not only from a technical perspective but also by improving emotional engagement between users and the model.

Many users have welcomed these changes. The combination of restoring GPT-4o, offering flexible model options, and increasing usage limits has been seen as strengthening the value of ChatGPT’s service. These updates also demonstrate that community feedback has a tangible impact on product development.

OpenAI emphasized that it will continue monitoring user reactions to these changes, including early trials of the new warmer GPT-5 personality, which is expected to roll out in the coming weeks.

With this round of updates, ChatGPT now offers a more balanced combination of speed, depth, and personalization. Users have more control to tailor the model to their needs, while OpenAI shows a clear commitment to listening and responding quickly to market feedback.

Subscribe to get the latest posts sent to your email.

Type your email…

Subscribe

ChatGPTGPTGPT-4oGPT-5OpenAI