OpenAI has unveiled its new reasoning-focused AI models, dubbed “o1-preview” and “o1-mini,” marking a significant advancement in AI capabilities. These models, part of the “Strawberry” family, are designed to enhance reasoning skills, particularly in math and science, and can fact-check themselves by thoroughly considering all parts of a command or question. Although these models cannot browse the web or analyze files yet and are currently rate-limited and expensive, OpenAI plans to make the o1-mini accessible to all free ChatGPT users in the future. The o1 models are available through ChatGPT and OpenAI’s API, showcasing improved performance in reasoning tasks compared to previous models.

<https://arstechnica.com/information-technology/2024/09/openais-new-reasoning-ai-models-are-here-o1-preview-and-o1-mini/>