OpenAI completes development of powerful AI model o3-mini with reasoning ability

OpenAI CEO Sam Altman announced on Platform X on Friday that work on the large o3-mini language model is complete and is expected to launch in the next couple of weeks. He added that the company has taken into account the wishes of users and plans to release the application programming interface (API) and ChatGPT at the same time.

Image source: Mariia Shalabaieva/unsplash.com

OpenAI unveiled next-generation large language models o3 and o3-mini with reasoning capabilities last December, describing them as the smartest AI solutions in the world.

The o3 and o3-mini models are superior in performance and capabilities to the “thinking” AI model o1, announced last September. The company previously announced that it plans to launch the o3-mini model by the end of January 2025, after which the full-scale o3 model will be released, hoping that more reliable and smart large language models will be able to outperform existing versions of neural networks, and this will attract new investments and increase the audience of users .

Also this week, OpenAI introduced a beta version of a new feature called Scheduled tasks in ChatGPT, allowing users to automate their work using scheduled tasks in ChatGPT. As Reuters noted, this indicates OpenAI’s “invasion” into the field of smart assistants, turning the ChatGPT chatbot into a competitor to personal assistants such as Siri, Alexa, etc.