Everything we know about GPT-5, OpenAI's new model that could be released "soon."

OpenAI is in luck, not only because of its 700 million weekly active users and the rollout of new features integrated into ChatGPT , but also because of the imminent release of the GPT-5 model .

Sam Altman, CEO of OpenAI, announced on his profile on the social network X (formerly Twitter) that OpenAI wants to "establish precise expectations" because it is an "experimental model that incorporates new research techniques" that they will use in future models. But what will it offer? Everything points to GPT-5 remembering people's most important tasks and things to become the best assistant on the market, making decisions efficiently, establishing more advanced interactions and being able to generate texts without making grammatical or spelling errors, among other new features.

However, beyond these capabilities, one of the great new features of GPT-5 is that it will be able to unify the functions of other conventional ChatGPT models with the capabilities of the O series models . This way, users will be able to switch models to use each artificial intelligence according to their needs.

On the other hand, GPT-5 will also enable more efficient and powerful capabilities in reasoning, conversation, multimodal tasks, and text generation, considering that this model could incorporate a context window of one million tokens . In addition, it will integrate the SORA tool for creating and processing video.

Also, The Verge reveals that the model will be released in mini and nano versions through its API . That is, the main version will combine a conversational large language model and a simulated reasoning model; however, the nano version will only be accessible through the API.

Finally, regarding its availability, Altman reveals in X that they will soon launch GPT-5 , although they have not yet announced a specific date.

New features coming to ChatGPT

ChatGPT incorporates reminders to send alert messages to users who have spent too much time with the chatbot . Thanks to this new feature, OpenAI encourages Internet users to take a break , taking into account that the alert messages arrive discreetly with this text: "You've been chatting for a while, does this seem like a good time for a break?" Additionally, once the alert appears, the chatbot offers two buttons: one to continue chatting and another to disconnect from the AI for a while.

OpenAI also acknowledges that it is working closely with experts to improve how ChatGPT responds in critical moments, such as when someone shows signs of mental or emotional distress .

And on the other hand, the latest innovation presented by OpenAI is related to personal challenges . This means that when someone asks something like "should I break up with my boyfriend?", the AI shouldn't provide an answer, but rather perform an analysis using questions to weigh the pros and cons .

Sign up for our newsletter and get the latest technology news straight to your inbox.