I built a local GPU benchmark server for 4B to 35B models because I was tired of synthetic cloud leaderboards and tiny sample tests

Hi everyone, whenever I looked at existing LLM leaderboards I was missing practical numbers for running quantized open weight models on affordable local hardware. Most tables only copy vendor cloud scores or run tiny subsets of a benchmark, and none of them tell you how long a model actually sits th

中文全文翻译正在补齐，可先查看下方英文原文。

Hi everyone, whenever I looked at existing LLM leaderboards I was missing practical numbers for running quantized open weight models on affordable local hardware. Most tables only copy vendor cloud scores or run tiny subsets of a benchmark, and none of them tell you how long a model actually sits th