OpenAI Developers(@OpenAIDevs)

Enter the AGI Era with GPT-6 Astra. GPT-6 Astra scores 75.2% on DeepSWE v1.1 at xhigh reasoning. A...

8.5内容质量
Enter the AGI Era with GPT-6 Astra. GPT-6 Astra scores 75.2% on DeepSWE v1.1 at xhigh reasoning. A...

TL;DR · AI 摘要

GPT-6 Astra在DeepSWE v1.1测试中达到75.2%，具备自动化修复bug的能力。

核心要点

  • GPT-6 Astra在DeepSWE v1.1 xhigh推理任务中得分75.2%
  • 支持跨文件追踪bug根源并生成修复方案
  • 可自动执行测试验证修复效果

结构提纲

按章节快速跳转。

  1. §AGI时代里程碑

    宣布GPT-6 Astra实现AGI关键突破

  2. ·基准测试表现

    在DeepSWE v1.1 xhigh任务中取得75.2%成绩

  3. ›智能调试功能

    支持五步自动化修复流程：复现-追踪-比较-补丁-验证

  4. ·技术特性

    跨文件代码分析与多方案比较能力

  5. ›应用价值

    显著提升软件缺陷修复效率

思维导图

用一张图看清主题之间的关系。

查看大纲文本（无障碍 / 无 JS 友好）
  • GPT-6 Astra的AGI时代
    • 技术突破
      • 75.2% DeepSWE得分
      • xhigh推理能力
    • 核心功能
      • 跨文件调试
      • 多方案比较
      • 自动测试验证

金句 / Highlights

值得收藏与分享的关键句。

#GPT-6 Astra#AGI#DeepSWE#OpenAI
打开原文

OpenAI Developers on X: "Enter the AGI Era with GPT-6 Astra. GPT-6 Astra scores 75.2% on DeepSWE v1.1 at xhigh reasoning. Ask Astra to fix a bug and have it: • Reproduce the bug from your report • Trace the cause across files • Compare fixes and side effects • Patch the code causing the failure … / X

OpenAI Developers

@OpenAIDevs

Enter the AGI Era with GPT-6 Astra. GPT-6 Astra scores 75.2% on DeepSWE v1.1 at xhigh reasoning. Ask Astra to fix a bug and have it: • Reproduce the bug from your report • Trace the cause across files • Compare fixes and side effects • Patch the code causing the failure • Run tests and retry the failing workflow

7:33 PM · Sep 3, 2026

30.9K

Views

18

22

452

44