OpenAI Launches o1 Reasoning Model, Touting Major Gains in Math and Science
Read the full article for context, quotes, and updates from the team.
OpenAI has unveiled o1, a new family of models built to tackle complex reasoning tasks in mathematics, coding and science. The company says the system is designed to think through problems step by step, using reinforcement learning to improve its ability to work through difficult questions before producing an answer.
According to OpenAI, o1 has delivered strong results on several challenging benchmarks. The model scored 83% on AIME 2024, a test focused on advanced math, and 74% on GPQA Diamond, a benchmark used to measure performance on graduate-level science questions. OpenAI says these results represent a significant improvement over earlier models, including GPT-4o, particularly on tasks that require careful reasoning rather than quick pattern matching.
The new model family is being positioned as a major step forward in AI systems that can handle more demanding academic and technical work. OpenAI said o1 is now available to ChatGPT Plus and Team users, with broader access planned in the near future.
The release underscores the company’s push to build models that do more than generate fluent text, instead focusing on deeper problem-solving capabilities that could be useful in research, software development and other high-skill fields.
