OpenAI has released its artificial intelligence model o1 from preview to all ChatGPT users.With both faster and more powerful thinking capabilities, o1 includes significant improvements in coding, math and writing.

OpenAI o1 stable version released

OpenAI stated that compared to the preview version, o1 is now able to produce more fluid and comprehensive responses. Accordingly, it will be able to solve complex problems from users in more detail. Perhaps the most striking innovation is o1’s ability to upload images.

Thanks to the support for uploading images, users will be able to send images to ChatGPT and receive more detailed answers. o1 model can analyze images, understand their content and answer questions using the information. On the other hand, it will also be able to annotate and produce creative content.

This feature expands the usage areas of ChatGPT, offering potential applications in many different fields from education to design. Unlike previous versions, the new model actually goes through more “thinking” process before answering the questions, in other words, it strategizes like a human.

The o1 family offers significant improvements especially in complex subjects such as math, physics, chemistry and coding. For example, in a recent test, the o1 model achieved an 83 percent success rate on the International Mathematics Olympiad exam.

The next-generation artificial intelligence model was asked several challenging questions from the International Mathematics Olympiad. OpenAI o1 solved 83 percent of these questions correctly, while GPT-4o only achieved a 13 percent success rate on the same questions.