OpenAI releases GPT-5.2 after “code red” Google threat alert

On Thursday, OpenAI launched GPT-5.2, its latest suite of AI models tailored for ChatGPT, featuring three distinct versions: Instant, Thinking, and Pro. This release comes on the heels of a critical 'code red' memo issued by CEO Sam Altman earlier this month, which aimed to redirect company efforts toward enhancing ChatGPT in light of rising competition from Google's Gemini 3 AI model. Fidji Simo, OpenAI’s chief product officer, emphasized during a press briefing that the new version is designed to unlock greater economic potential for users. 'GPT-5.2 is better equipped for tasks such as creating spreadsheets, building presentations, writing code, interpreting images, and managing complex, multi-step projects,' Simo explained. The three tiers of GPT-5.2 cater to varying user needs. Instant is optimized for rapid tasks like writing and translation, while Thinking is geared towards simulated reasoning for more intricate challenges such as coding and mathematics. Pro aims for the pinnacle of performance, providing high-accuracy responses for particularly tough problems. Notably, GPT-5.2 boasts a substantial 400,000-token context window, allowing it to process extensive amounts of information simultaneously. Its knowledge cutoff is set for August 31, 2025. The rollout for paid ChatGPT subscribers began Thursday, with API access available to developers at a cost of $1.75 per million input tokens for the standard model—a 40% increase over the previous version, GPT-5.1. OpenAI has confirmed that GPT-5.1 will continue to be accessible to paid users for three months via a legacy models dropdown. This latest release follows a challenging period for OpenAI, as Altman’s internal 'code red' directive was prompted by Google’s Gemini 3 model surpassing multiple AI benchmarks and capturing market share. The memo called for a pause on other initiatives, such as advertising plans for ChatGPT, to concentrate on refining the chatbot’s core functionality. The stakes for OpenAI are considerable, with commitments amounting to $1.4 trillion for AI infrastructure development over the coming years, which were made when the company was perceived to have a clearer technological advantage. Currently, Google's Gemini boasts over 650 million monthly active users, while OpenAI claims 800 million weekly active users for ChatGPT.