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

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

The rave reviews OpenAI's latest models have been winning come with an asterisk: Experts are also finding that they're ...

In training the models, OpenAI used “reinforcement learning”. How does this method help AI models make better decisions when ...

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

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

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

OpenAI's o3 and o4-mini models for ChatGPT have arrived.

OpenAI launches groundbreaking o3 and o4-mini AI models that can manipulate and reason with images, representing a major ...

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

Following the recent launch of a new family of GPT-4.1 models, OpenAI released o3 and o4-mini on Wednesday, the latest ...

OpenAI touts o3 as a smart AI model with the ability to reason (meaning it can recursively check its answers before giving ...