Podcast claims OpenAI's internal multi-agent LLM cracked Navier-Stokes with ~10,000 agents

FlorianGallwitz · x · 2026-09-21

A German-language post points to an interview with Noam Brown discussing a new, still-internal multi-agent LLM at OpenAI, claimed to have recently solved the Navier-Stokes Millennium Prize problem using roughly 10,000 agents. The claim comes via podcast discussion and is unverified by OpenAI, so treat with caution.

Related event: Report: OpenAI Used ~10,000 Agents to Tackle Navier–Stokes(2 posts)→

Original post →

More from Models

Models channel →