OpenAI has launched o3-mini, a new AI reasoning model designed for STEM fields, offering improved speed and affordability. It makes 39% fewer major mistakes than o1-mini in real-world tests. While not OpenAI’s most powerful model, o3-mini provides a specialized alternative for programming, math, and science. It is available via ChatGPT and OpenAI’s API but does not yet support image analysis.
Source – CGTN