ChatGPT Gets a Powerful Model with a Free Version — o3-mini

OpenAI has released an even more powerful model, ChatGPT, compared to its predecessors. It also received a free version for review.

According to The Verge, the o3-mini model works 24% faster than o1-mini, while providing more accurate answers. In addition, the neural network shows the entire logical chain of reasoning, explaining how the result was obtained. The model is able to bypass its competitors in the field of writing code and logic with less effort. You can try it out for free — to do this, you need to select the function called Reason in ChatGPT.

Recall that the first results of testing o3 appeared in December. The neural network demonstrated an almost human level of reasoning — a number of experts even announced the imminent appearance of self-learning AI.

Microsoft has already added o3-mini to some of its products, including Microsoft Azure OpenAI Service and GitHub Copilot.

Earlier, a new model of the Chinese neural network DeepSeek was released, which brought down NVIDIA shares. Microsoft and OpenAI suspected that ChatGPT data was used for its training. We talked about the situation and AI capabilities in more detail in this article.