OpenAI delivered advanced ChatGPT reasoning models this month that are more capable than o1, but they also hallucinate more.

The FrontierMath benchmark from Epoch AI tests generative models on difficult math problems. Find out how OpenAI’s o3 and ...

A discrepancy between first- and third-party benchmark results for OpenAI's o3 AI model is raising questions about the ...

By OpenAI 's own testing, its newest reasoning models, o3 and o4 -mini, hallucinate significantly higher than o1.

OpenAI ChatGPT o3 and o4-mini models can determine location information by looking at a photo - and yes, it can be fooled.

OpenAI’s newest LLM, o3, is facing scrutiny after independent tests found it solved a far fewer number of tough math problems ...

OpenAI released upgraded versions of its advanced reasoning models. These new models, named o3 and o4-mini, offer ...

Historically, each new generation of OpenAI's models has delivered incremental improvements in factual accuracy, with ...

Learn how OpenAI's o3 and o4 models are setting new standards in generative AI, empowering businesses, developers, and ...

In December 2024, OpenAI held a livestream on YouTube and other social media platforms, announcing the o3 AI model. At the ...

According to internal tests, newer models like o3 and o4-mini hallucinate significantly more than older versions, and OpenAI doesn't know why.

OpenAI’s o3 model shows inflated benchmark results; real-world tests reflect performance far below initial FrontierMath ...