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.<p>Method: it takes the patches that benchmarks already publish (DeepSWE, SWE-bench Verified, SWE-bench Live) and counts "coding units": syntax elements like statements, calls, names and literals. I didn't use lines of code, so cramming code isn't rewarded. Deleted code is subtracted, so it measures net units.<p>GPT-6 and Claude 5.5 have no public patches yet, so I ran them myself on 3–10 tasks on my own subscriptions (shown as faded circles, not ranked).<p>Any feedback or model contributions are welcome, especially public patch exports for newer models.<p>GitHub: <a href="https://github.com/La5u/Parsimony" rel="nofollow">https://github.com/La5u/Parsimony</a><p>Not affiliated with any of the benchmarks.