GPT-5 premieres three modes of use and multiplies by 15 its weekly messages of messages for chatgpt plus users

GPT-5 has not taken in receive your first face washing important since it came to light. Just a week after its launch, Openai has introduced a change that directly affects how we interact with the model within Chatgpt: now You can choose between several operating modes that adjust their “way of thinking” in each answer.

Sam AltmanCEO of the company, announced it in its X (formerly Twitter) account early in the day. From this moment, users can alternate between three options: Auto, Fast and Thinking. The idea is simple, but the experience completely changes: “Auto” lets GPT-5 himself decide how long to invest in generating the answer; “FAST” prioritizes immediacy and “Thinking” is the one that takes the longest to elaborate more complete and with the highest level of detail.

This movement does not arise by chance. The launch of GPT-5 was accompanied by criticism about its performance, especially since Openai had reduced the variety of models and eliminated options that many users considered essential. With the introduction of these modes, The company intends to give more control to the community and offer different levels of reasoning According to the need for each moment.

A step back to take impulse

It is not the first time that Openai rectifies in a few days. After complaints about the disappearance of previous models, the company recovered the GPT-4O version for payment subscribers. And now, in addition to returning variety, adds the possibility of customizing the “speed of thought” in GPT-5. This change seems to be part of a broader strategy to recover the confidence of the most demanding users.

But not everything stays there. Together with the new modes, OpenAi has decided to relax one of the most controversial limitations of the new model: The weekly messages of messages.

From 200 to 3,000 messages per week

At its premiere, GPT-5 thought mode was limited to 200 weekly messages for users of Chatgpt plusa number that many considered ridiculous, taking into account that they pay 23 euros per month for the service. Altman has confirmed that This roof rises to 3,000 weekly messagesa huge leap that should leave much more margin to work without being aware of interactions.

In addition, the CEO recalled that the context limit of the Thinking mode stands at 196,000 tokenswhich allows you to keep very long conversations and tasks without losing information. Of course, it has also nuanced that these limits could change in the future based on the real use that is made of the service.

The chatgpt-5 pro enigma

A few days ago, Altman hinted that Chatgpt-5 Pro could reach the subscribers of Plus. However, It seems that this idea is parked at the moment. The reason is simple: “We do not have computer resources to do it right now”, Recognized the manager. This suggests that, although OpenAi’s roadmap includes new versions and services, we will have to wait for the infrastructure to assume that extra load.

In just one week, GPT-5 has gone from being Openai’s most anticipated model to star in a start with rapid criticism and rectifications. The company is moving rapidly to correct perception and functionality errorsintroducing improvements that, although they may seem small, have a real impact on the user experience.

The new ways allow to adapt the behavior of the model to each situationfrom obtaining an immediate response to something urgent until the necessary time is taken for a deep analysis. The increase in the weekly messages, on the other hand, eliminates one of the greatest frustrations of the payment subscribers.

It remains to be seen if these changes will be enough to convince those who felt disappointed in the first days. What is clear is that Openai is listening and reacting quickly, aware that in the competitive world of generative artificial intelligence, user’s confidence is as valuable as the power of the model itself.