OpenAI unveils o1 artificial intelligence with reasoning capabilities

OpenAI has unveiled its new artificial intelligence model called o1, which is capable of reasoning and solving complex problems.

This model, previously codenamed Strawberry, is capable of performing more complex tasks in science, programming, and mathematics than previous AI models. In addition to the basic model of o1, which is called o1-preview, OpenAI has also unveiled a more affordable version of o1-mini. These models are trained to spend more time thinking about tasks before responding, allowing them to improve their strategies and detect errors.

In the tests conducted, the o1 artificial intelligence model competed with master’s students in physics, chemistry and biology and was able to get high marks in mathematics and programming. In the preliminary exam of the International Mathematical Olympiad, o1 was able to solve 83% of the problems correctly, while the previous model GPT-4o solved only 13% of the problems. Even though the o1 model still does not have many features such as web browsing and file upload, having the ability to reason and solve complex problems is considered a step forward in the field of artificial intelligence.

OpenAI has made the o1 artificial intelligence model available to ChatGPT Plus and Team users with a limit of 30 messages per week for o1-preview and 50 messages for o1-mini. The company plans to provide o1-mini to all ChatGPT users in the future.