OpenAI releases GPT-5.2 after Sam Altman's ‘Code Red’—what's new?

In a bold move just a month after its previous ChatGPT update, OpenAI has launched GPT-5.2, coinciding with CEO Sam Altman's declaration of a 'Code Red' in response to the fierce rivalry posed by Google's Gemini 3. The company announced that GPT-5.2 is the most advanced model designed for professional tasks yet. As stated in their release, this iteration significantly enhances capabilities in general intelligence, long-context comprehension, and tool utilization, thereby excelling in executing complex tasks more efficiently than prior versions. Despite the timing of this release, OpenAI clarified that GPT-5.2 is not simply a reaction to the launch of Gemini 3. Reports had hinted at a potential release due to the heightened competitive atmosphere. OpenAI noted that heavy users of ChatGPT Enterprise have reported saving up to 10 hours weekly, and GPT-5.2 aims to further boost productivity compared to its predecessor, GPT-5.1. This new model promises to outperform previous versions in various applications, including creating spreadsheets, generating presentations, coding, interpreting images, and managing intricate, multi-step processes. Feedback from early users has been overwhelmingly positive, with industry leaders like Jeff Wang, CEO of Windsurf, heralding GPT-5.2 as a revolutionary advancement in coding capabilities. Additionally, OpenAI revealed that GPT-5.2 has achieved new benchmarks in abstract reasoning, demonstrating significant improvements over GPT-5.1. The new features, including GPT-5.2 Instant, Thinking, and Pro, began rolling out in ChatGPT for paid subscriptions on the day of the announcement. In a surprising turn, during a CNBC interview, Altman stated that the launch of Gemini 3 had less impact than he initially expected. He emphasized the importance of addressing competitive threats swiftly and expressed confidence that OpenAI would overcome its 'Code Red' status by the end of January 2022. Meanwhile, Google has touted Gemini 3's superior performance across various benchmarks, claiming it marks a significant advancement toward achieving artificial general intelligence (AGI).