GPT-5 Pro, GPT-5 Thinking, and o3 to analyze my code repository - and found surprising differences in detail, reasoning, and ...

OpenAI has since posted some updated charts on its website. The new deception rate chart certainly suggests that a mere mistake was made. The revised stats show GPT-5's coding deception rate at 16.5%, ...

Following this post’s publication, OpenAI co-founder and CEO Sam Altman announced the company would restore access to GPT-4o and other old m ...

Specifically, GPT makes incorrect claims 9.6 percent of the time, compared to 12.9 percent for GPT-4o. And according to the ...

OpenAI’s o3 has defeated xAI’s Grok in a chess match that took place on Google’s Kaggle platform. But what does it prove?

Google's AI chess tournament ended Thursday with OpenAI’s o3 model sweeping xAI's Grok in four straight games.

Some charts featured in its OpenAI's GPT-5 livestream on Thursday had several mistakes, which CEO Sam Altman called a "mega ...

Openai is making O3 level capabilities open source ahead of the release of GPT-5 in about two days. These open source models ...

The o3 model also failed to solve more than 100 visual puzzle tasks, even when OpenAI applied a very large amount of computing power toward the unofficial score, said Mike Knoop, an ARC Challenge ...

The o3 model scored 76% accuracy on ARC-AGI in an evaluation formally coordinated by OpenAI and the author of ARC-AGI, François Chollet, a scientist in Google's artificial intelligence unit.

Elon Musk's Grok 4 was the strongest fighter in the eight-player field until the final. But in the final, it made ...

o3-mini is free for all When OpenAI announced the new model in December, my first thought was "how long will we have to wait to try this for free?" Luckily, Altman and co has given free users a ...