OpenAI introduces its most advanced AI model, o3-pro

OpenAI, one of the leading companies in the artificial intelligence sector, has announced its latest AI model, o3-pro. This new version, which is an improved iteration of the o3 model released earlier this year, stands out as the most powerful AI model the company has developed to date, according to OpenAI's statement.

OPENAI O3-PRO INTRODUCED

The o3 series, which OpenAI describes as a "reasoning model," distinguishes itself from classic language models by having the ability to solve complex problems step-by-step. This feature makes it more reliable, especially in fields where accuracy and process management are critical, such as physics, mathematics, and programming.

AVAILABLE TO CHATGPT PRO AND API USERS

Starting today, o3-pro has been made available to ChatGPT Pro and Team subscribers. Enterprise and Edu users will gain access to this model next week. For developers, the model has been made accessible via API and operates on a paid system: the API charges 20 dollars per million input tokens and 80 dollars per million output tokens. For comparison, one million tokens correspond to approximately 750 thousand words; this means a text slightly longer than Tolstoy's "War and Peace." With this pricing, o3-pro has a cost 10 times higher than the o3 model.

According to OpenAI's statement, the o3-pro model can search the web, examine visual content, analyze files, execute Python code, and utilize its memory to generate personalized responses. However, these advanced features mean a slight increase in response time; o3-pro responds slightly slower than its predecessor, o1-pro.

In addition, the model has some limitations. Currently, the temporary chat feature is disabled in ChatGPT. Furthermore, o3-pro cannot generate visual content and is not yet compatible with Canvas, OpenAI's AI-powered collaboration platform.

Source: donanimhaber.com

News Source: 12punto