OpenAI has announced the release of two new AI models, o3 and o4-mini, introducing the ability to process and reason with visual information, a feature that allows them to interpret images such as sketches and whiteboards.
The o3 model is described as OpenAI’s most powerful reasoning model to date, capable of handling complex tasks that require advanced problem-solving skills. Meanwhile, o4-mini offers a more compact and efficient alternative, delivering impressive performance in areas such as mathematics, coding, and visual analysis at a lower cost.
Both models have been designed to utilize the full suite of tools available in ChatGPT, including web browsing, Python execution, image generation, and file interpretation.
This integration allows for more dynamic and context-aware interactions, enhancing the user experience across various applications. These tools are now accessible to ChatGPT Plus, Pro, and Team users for o3, o4-mini, and o4-mini-high models, with plans to extend availability to o3-pro in the coming weeks.
The release of these models follows OpenAI’s recent launch of GPT-4.1, a flagship AI model that boasts a significantly larger context window, supporting up to one million tokens. GPT-4.1 has demonstrated notable improvements in coding and instruction-following capabilities.
Introducing OpenAI o3 and o4-mini—our smartest and most capable models to date.
For the first time, our reasoning models can agentically use and combine every tool within ChatGPT, including web search, Python, image analysis, file interpretation, and image generation. pic.twitter.com/rDaqV0x0wE
— OpenAI (@OpenAI)
In addition to technological advancements, OpenAI has updated its preparedness framework for evaluating the risks posed by new AI models. The revised framework introduces new research categories to address emerging threats, such as an AI model’s ability to replicate itself, conceal its capabilities, evade safeguards, or prevent shutdowns.
