Last December, Openai presented 03-mini, a new Advanced reasoning model This, at some point in the future, would happen to Chatgpt. And this moment is now. The company led by Sam Altman has already published the new 03-mini for all users.
This model allows the chatbot respond and solve the most complex problems in -depthSomething that can be very useful in the situations in which we are looking for a much more elaborate, precise and detailed response or an explanation. You can start using 03-mini, even if your account is free.
The new O3-Mini is available for free (but limited)
Openai presented GPT 01 in September 2024, as a model capable of “thinking better” your answers, even if you take more time to answer our questions. This model is focused on reasoning, so I would also be able to recognize and understand their mistakes, in case of cases. The big difference in GPT O1, in substance, is that instead of trying to offer us an answer as quickly as possible (something that the chatbot usually does), Your priority is to offer us the best answer.
Thanks to this reasoning capacity, the model can Think and analyze before giving an answer, just like a personSaid Openai at that time. Now, with 03-mini, the capacities of this model have improved, and Openai has already published it for everyone. According to the company, the O3-Mini tests have shown that it had reduced errors in the most difficult issues by 39%, against 01-mini.
How can we use O3-Mini? In the web version of Chatgpt, we will now see a new button called “reason” (or “reason”, translated into Spanish), and by pressing it, we will indicate to the chatbot that we want to reason with their response; To do this, you will use the 03-Mini model. As we explain, in this way, Chatgpt could take a little more time to offer us the answer, but it will be a much more detailed and developed answer.
This model is ideal for answering Consultations linked to science, technology, engineering and mathematics. On the other hand, for tasks such as writing an email, a presentation letter or similar, it could be a better option to use the standard Chatgpt model.
The new model of reasoning O3-Mini It is now available for all chatgpt users From the web version, even for free accounts. However, if you have a free account, use is limited. Among the others, Chatgpt Plus users will be able to send up to 150 messages per day using the reasoning model, while chatgpt pro users can use the 03-Mini model in an unlimited manner.
Cover photo | Sanket Mishra + Openai (with edition)
In Xataka Android | How to chat with chatgpt on WhatsApp without having to record a phone number on the agenda
In Xataka Android | Battle of IAS. I compared GPT-4O with Gemini 1.5 pro and I am clear that the fight is more suitable than ever
