%20(1).png)
OpenAI said on Friday that it was testing two reasoning AI models, o3 and o3 small, as part of a growing race with rivals like Google to develop better models capable of addressing complicated challenges.
CEO Sam Altman stated that the AI firm intends to unveil o3 small by the end of January and full o3 after that, since more robust big language models may surpass existing models and attract additional investors and customers.
In September, Microsoft-backed OpenAI introduced o1 AI models, which are supposed to spend more time processing queries and solving difficult issues.
The o1 models can reason through complicated tasks and tackle more difficult challenges than earlier models in science, coding, and arithmetic, according to the AI firm's blog post.
OpenAI's new o3 and o3 small models, which are now undergoing internal safety testing, are expected to be more powerful than the company's previously released o1 versions.
The GenAI pioneer said that it was starting an application procedure for other researchers to test o3 models prior to the public release, which will end on January 10.
OpenAI sparked an AI arms race when it introduced ChatGPT in November 2022. The company's rising popularity and new product releases helped OpenAI close a $6.6 billion fundraising round in October.
Rival Alphabet's Google published the second generation of its AI model Gemini earlier in December, as the search giant seeks to recover the lead in the AI technological race.
