Image understanding is powered by multimodal GPT-3.5 and GPT-4. “These models apply their language reasoning skills to a wide range of images, such as photographs, screenshots, and documents containing both text and images.”

→ View original post on X — @officiallogank,