OpenAI推出新一代大模型GPT-5.2

Core Insights - OpenAI has officially released its new model GPT-5.2, focusing on general intelligence, coding capabilities, and long-context task processing, following Google's Gemini 3 update [3][5]. Product Performance and Structure - GPT-5.2 has improved speed in information retrieval, writing, translation, mathematical reasoning, and programming tasks, and can handle longer and more complex reasoning chains [5]. - The model is divided into three versions: Instant, Thinking, and Pro, catering to different needs such as instant responses, deep reasoning, and high-performance tasks [5][9]. - Pricing for the API has been adjusted, with input costs at $1.75 per million tokens and output costs at $14 per million tokens, with a 90% discount on cached input [5]. Competitive Landscape - OpenAI's recent updates come after mixed feedback on GPT-5 and a notable performance from Google's Gemini 3 in various industry tests, prompting OpenAI to accelerate product iterations [7]. - OpenAI's CEO stated that the impact of Gemini 3 on internal metrics was less than expected, with plans to end the "red code" status by January [7]. User Management and Content Licensing - Disney has announced a $1 billion investment in OpenAI, allowing the use of its characters in OpenAI's Sora video generation tool [9]. - OpenAI is enhancing user management capabilities, including testing age verification features to implement content and model access restrictions [9]. - The three-tier structure of GPT-5.2 aligns with the company's strategy to offer tailored capabilities for different user needs, with plans for additional tools for enterprise content automation and permission management [9].