OpenAI’s o3 and o4-mini: Smarter Models with Tool Use and Visual Reasoning

OpenAI has introduced two new models, o3 and o4-mini, as part of its o-series, showcasing significant advancements in autonomous tool use and visual reasoning. These models can independently decide when and how to use ChatGPT’s integrated tools—such as web browsing, Python for data analysis, and image handling—to solve complex tasks, often in under a minute. A standout feature is their ability to manipulate images during reasoning, like rotating and transcribing upside-down handwriting. Both models are now available to paying ChatGPT users, with a more powerful o3-pro version expected soon.