---
id: 20261014-T0-09
title: "Show HN: Parsimony – an open-source benchmark for how much code AI models add"
url: https://ai.daily.yangsir.net/daily/20261014-T0-09
issue_date: 2026-10-14
publish_date: 2026-10-13T19:50:31.000Z
source_name: "HN AI 精选"
source_url: https://parsimony.lasu.dev/
---

# Show HN: Parsimony – an open-source benchmark for how much code AI models add

Existing benchmarks measure solve rate and token output/cost, but don't punish excess code, which makes a codebase harder to work with in the long term. I wanted to measure this since it matters most for long-term projects. Method: it takes the patches that benchmarks already publish (DeepSWE, SWE-b

---

**来源**：[HN AI 精选](https://parsimony.lasu.dev/)

**详情页**：https://ai.daily.yangsir.net/daily/20261014-T0-09

---

*智语观潮 · Daily — https://ai.daily.yangsir.net/llms.txt*