# Rohan Paul 评 OpenAI 未发布内部模型能力曲线显著高于 GPT-6 Astra

- 来源：Rohan Paul (@rohanpaul_ai)
- 发布时间：2026-09-09 03:01
- AIHOT 分数：66
- AIHOT 链接：https://aihot.news/items/cmtt1o7w5014srowa5qzuspmb
- 原文链接：https://x.com/rohanpaul_ai/status/2097400068169253143

## AI 摘要

Rohan Paul 转述 OpenAI 官方消息：其未发布内部模型在多项基准上相比 GPT-6 Astra 有阶跃式提升，训练仍在进行。该模型组用约 10,000 个协作 AI 智能体在 88 小时内得出 Navier–Stokes 解，全程保持监测与隔离等前沿评估安全防护。作者认为新模型能更高效地利用额外推理计算，而非简单延长 Astra 的运行时间。

## 正文

OpenAI's unreleased internal model sits on a much higher capability curve than GPT-6 Astra.

the one that just solved the 90-year old Navier-Stokes Millennium Prize math Problem.

the new model also appears to use additional reasoning compute much more productively. That's how throwing 10,000 agents and enormous compute at Navier-Stokes could produce something qualitatively different from simply running Astra longer.

### 引用推文

> OpenAI：This model represents a step-function improvement on many benchmarks, and its training is ongoing. Our internal model group arrived at the Navier–Stokes solutio...