**Benchmarks show scores. Dashboards show usage. What ships?**

> Benchmarks rank models. Usage dashboards rank adoption. Pragmatikos measures what developers feel: how many turns it takes to ship, which planner-builder pair gets there, and what a delivered outcome costs. For OpenCode 2.0 Benchmarks test single model. Developers often pair them. A benchmark is an exam: a clean task, a hidden answer key, one model, nobody steering. Real work is the job: a messy repo, your tools, you...

**来源信息**
- **来源**：Hacker News：AI 热帖
- **分类**：论文
- **发布时间**：2026-09-08 00:16（北京时间）
- **原文**：[打开原文](https://pragmatikos.ai/)