OpenAI unveils updated version of ChatGPT

US-based OpenAI has released its new AI model, GPT-5.2, optimised for knowledge workers. The new model performs better in mathematics, solving scientific problems and programming, OpenAI says.

“We designed 5.2 to unlock even more economic value for people,” OpenAI’s Head of Applications Fidji Simo told journalists. “It’s better at creating spreadsheets, building presentations, writing code, perceiving images, understanding long context, using tools and then linking complex, multi-step projects.”

About a week ago, OpenAI Chief Executive Officer Sam Altman declared a “code red” within the company to improve the quality of ChatGPT, postponing the development of other projects, including those in advertising.

The company stepped up its efforts after Google released a new version of Gemini that outperformed ChatGPT across a number of benchmarks, including expert knowledge, logical puzzles, mathematical tasks and pattern recognition.

In addition, OpenAI is facing growing pressure in the business segment from Anthropic, which has recently strengthened its work in this area.

GPT-5.2 has been optimised in terms of reasoning ability, programming, and handling a wide range of inputs, including text, audio and video — areas where OpenAI has faced challenges from Google and Anthropic, said Gartner analyst Arun Chandrasekaran, as quoted by The Wall Street Journal.

OpenAI has also made progress in the cognitive work functions required for professional workers to choose ChatGPT as their AI assistant, he added.

By Vugar Khalilov