OpenAI Unveils Advanced AI Model Capable of Analyzing Diagrams and Sketches

OpenAI has launched its latest AI model, o3, claiming it can understand and analyze visual information such as diagrams, whiteboards, and sketches—even if they are low quality. The company also introduced a smaller, faster o4-mini model. Both are designed to integrate visual data into their reasoning processes, representing a significant advance from previous models that focused predominantly on text. OpenAI’s goal is to keep ahead of rivals like Google, Anthropic, and xAI in the increasingly competitive generative AI sector. The new models are especially capable in math, coding, science, and image comprehension, and have undergone rigorous safety testing. However, recent changes to OpenAI’s safety policies and the discontinuation of some transparency measures have drawn scrutiny from observers.

Read more