OpenAI claims Navier-Stokes Millennium Prize solution from agent swarm on next-gen model

RexDouglass · x · 2026-09-09

OpenAI announced that a group of agents, powered by a next-generation model it says is significantly more capable than GPT-6 Astra, produced a solution to the Navier-Stokes Millennium Prize Problem — whether smooth 3D fluid motion can break down — open for roughly 90 years.

The timeline runs from Navier (1822) and Stokes (1845) through Leray's weak solutions (1934), Clay's $1M prize (2000), numerical Euler blowup (2014), Tao's averaged blowup proof (2016), Elgindi's rough-data Euler blowup (2021), and forced blowups by Córdoba & Martínez-Zoroa. This August, Buckmaster and Alpöge obtained smooth-forced blowup for Boussinesq and Euler with Claude and Codex, verified in Lean, before OpenAI announced the agent-produced proof of the full problem.

Related event: OpenAI Claims Navier-Stokes Millennium Solution; Verification and Attribution Disputes Erupt(249 posts)→

Original post →

More from Models

Models channel →