OpenAI launches GPT-5.2 on the move in ‘Code Red’

I introduced OpenAI GPT-5.2, the smartest AI model to date, with performance gains across typing, coding, and inference benchmarks. The launch comes just days after CEO Sam Altman internally announced “Code Red,” a company-wide campaign to improve ChatGPT amid intense competition from competitors.

“We announced this icon red to really signal to the company that we want to organize resources in a certain area, and this is a way to really prioritize,” Fidji Simo, chief applications officer at OpenAI, said in a press conference on Thursday. “We’ve had an increase in resources focused on ChatGPT in general.”

Simo denied that OpenAI had bumped up the release of GPT-5.2 in light of its code red, claiming that the company had been working on releasing that model for months. However, she said the additional resources about ChatGPT were “helpful.”

While OpenAI’s models and products were considered best-in-class when ChatGPT launched in 2022, this is no longer a foregone conclusion. The startup now faces a host of credible competitors, perhaps none more threatening than the recently launched Google Gemini models 3 It was well received by the technology industry. Google’s Gemini app has grown at an impressive rate over the past year, now boasting more than 650 million monthly active users, compared to OpenAI’s 800 million weekly active users. This pressure has forced OpenAI to rein in some of its most ambitious projects, including its work on information delivery Advertisements on ChatGPT, And refocus on improving core technology and products.

Just as the company’s recent models launched, GPT-5.2 ships as a series of models: Instant, which responds faster and is better for finding information; Thinking that excels in programming, mathematics, and planning; and Pro, the most powerful level in OpenAI models that provides higher accuracy on difficult questions.

OpenAI calls GPT-5.2 its best model yet for everyday professional use. GPT-5.2 Thinking achieved the highest scores to date on the GDPR test (GDVval), an OpenAI benchmark that compares performance between AI models and human professionals across 44 real-world professions. The company says the model outperforms human professionals on more than 70% of tasks, completing them 11 times faster.

The new version should also offer a significant reduction in hallucinations, says Max Schwarzer, post-training lead at OpenAI. The company says GPT-5.2 thinking causes 38 percent less hallucinations than GPT-5.1 on benchmarks measuring answers to factual questions.

The company offers GPT-5.2 to both ChatGPT users and developers on OpenAI’s API product. OpenAI says the new series of models “deliver clear gains across everyday and advanced use cases.”