OpenAI releases GPT-5.2 after triggering ‘red alert’

OpenAI claims GPT-5.2 is the most powerful AI model ever for everyday professional use, outperforming its predecessors in creating spreadsheets, building presentations, recognizing images, writing code, and understanding long-form contexts. The model is now integrated into ChatGPT and its application programming interface (API).

Previously in November, OpenAI released the GPT-5.1 update. However, two competitors Google and Anthropic also released new models that were highly appreciated by the technology community, causing OpenAI to declare a “state of emergency” to improve ChatGPT and temporarily shelve other projects.

“We activated the red alert to emphasize that the company needs to put resources into a certain area, clearly identify priorities and things that can be de-prioritized,” Fidji Simo, CEO of applications at OpenAI, confirmed on December 11. “We increased resources for ChatGPT. I think that helped with the release of the new model, but it’s not the main reason why it launched this week.”

Sam Altman, CEO of OpenAI, said Google’s release of Gemini 3 affected the company’s metrics less than initially feared. He expects OpenAI to escape the red alert status in January 2026. “When a competitive threat arises, you need to focus and resolve it quickly,” Altman told CNBC.

GPT-5.2 will have an Instant version – optimized for speed for common queries such as searching for information, writing, translating, Thinking – excelling in complex structured work such as programming, analyzing long documents, calculating, planning, and Pro – the most advanced model to bring maximum reliability and accuracy to difficult problems.

According to OpenAI, GPT-5.2 ranks top in many tests, including SWE-Bench Pro for programming performance, and GPQA Diamond for graduate-level scientific reasoning. On GDPval, a test published by OpenAI earlier this year, the new model outperformed or equaled industry leaders in 70.9% of well-described tasks.

TechCrunch Citing a source, image creation will be OpenAI’s top priority in the near future, especially after Google’s Nano Banana Pro makes a big splash. The company also plans to release another model next month with higher picture quality, faster speed and better personality.