OpenAI officially releases the new generation large language model GPT-5.2 today, marking the company's further acceleration in expanding into enterprise application scenarios, upgrading ChatGPT from a consumer tool to an enterprise productivity engine. The new model emphasizes faster speed, greater stability, and the ability to handle complex tasks, integrate multi-step workflows, and execute professional workstreams. With major partners such as the US government and Disney joining, OpenAI's business layout is rapidly expanding.
OpenAI states that the core goal of GPT-5.2 is to "create greater economic value," with significant enhancements in spreadsheet processing, presentation generation, code writing, visual recognition, long-context understanding, and tool invocation. To validate these capabilities, the company has introduced a proprietary evaluation system called GDPval, which simulates 44 different professional tasks to assess model performance. According to reports, GPT-5.2 achieves or surpasses human professional levels in approximately 71% of tests. While the external community is still awaiting independent review, this data has garnered widespread industry attention.