OpenAI has unveiled a new o1 artificial intelligence model that can reason and solve complex scientific problems

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

Here's What We Know

The model, previously known by the codename Strawberry, can handle more complex tasks in science, programming and maths compared to previous AI models. In addition to the basic o1 model, the company has also introduced a more affordable o1-mini version. These models are trained to spend more time thinking about tasks before answering, allowing them to improve their strategies and recognise errors.

In the tests, the o1 model performed comparable to postgraduate students in physics, chemistry and biology, and also achieved high scores in maths and programming. In the qualifying exam for the International Mathematical Olympiad, the o1 model solved 83% of the problems, whereas the previous GPT-4o model only managed 13%. Although the o1 model does not yet have many features such as web browsing and file uploading, it represents a significant step forward in AI.

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

Source: OpenAI