[OpenAI Internal Model Solves Millennium Math Problem in 88 Hours] DeepTech TechFlow reports that on September 9, citing Jin10 Data, OpenAI announced on Tuesday local time that its next-generation internal model successfully solved a Millennium Prize Problem related to the Navier-Stokes equations. The model achieved a pass rate of nearly 48% on selected open mathematical problems, approximately three times the peak performance of GPT-6 Astra. The research team deployed around 10,000 collaborative AI agents, arriving at the solution approximately 88 hours after the initial launch, with total computational costs amounting to several million dollars. The possibility of singularities forming in three-dimensional fluid motion within finite time could lead to the breakdown of existing equation descriptions. New York University mathematician Tristan Buckmaster and Anthropic researcher Levent Alpöge had recently published related AI-assisted research. Buckmaster questioned whether OpenAI adopted a rare method, which took months to develop, after learning of his research. OpenAI responded, stating that their team had not accessed Buckmaster's work prior to its publication and that the proof methods used by both parties showed significant differences.