The upgraded O3 model from Openai can use images when thinking

Openai releases two new models to think of artificial intelligence: O3, which the company calls “the most powerful thinking model”, and O4-MINI, a smaller and fastest model “that achieves a remarkable performance of its size and cost”, ” According to the publication of the blog.

The company also says that O3 and O4-MINI will be able to do so Think with picturesThis means that they “will merge the pictures directly in their ideas series.” This may be useful if models such as graphics or white panels are displayed. Openai says the models will also be able to adjust images by zooming on them or rotate the image “as part of the thinking process.”

In addition, Openai announces that their thinking models will be able to use all ChatGPT tools, including things like web browsing and images. The tools will be available today for Chatgpt Plus, Pro and Team users in O3, O4-MINI and O4-MINI, and will come to O3-PRO in “a few weeks”. (Gradually get rid of O1, O3-MINI and O3-MINI from those levels.)

Today’s ads followed Openai on the pioneering GPT-4.1 AI model on mondayBehind it for GPT-4O.

Update, April 16X Patrosion of Openai.