# AI 可胜任多大规模的软件工程任务？

- 来源：Epoch AI (@EpochAIResearch)
- 发布时间：2026-04-10 23:22
- AIHOT 链接：https://aihot.virxact.com/items/cmnw1yqia00rdslc3qgdo9nnb
- 原文链接：https://x.com/EpochAIResearch/status/2042624189421752346

## AI 摘要

新基准 MirrorCode 显示，Claude Opus 4.6 能重构 16,000 行生物信息学工具包，任务量相当于人类工程师数周工作。与 METR_Evals 合作开发。

## 正文

What are the largest software engineering tasks AI can perform？

In our new benchmark， MirrorCode， Claude Opus 4.6 reimplemented a 16，000-line bioinformatics toolkit - a task we believe would take a human engineer weeks.

Co-developed with @METR_Evals. Details in thread.
