**Show HN: Sipp – Run small local LLMs in browser 3x faster**

> Sipp is a zero-dependency AI runtime with the fastest WebGPU backend on the web. Run models in the browser, add a cloud gateway, or go native on CUDA, Vulkan, and Metal. Sipp Pink Lemonade Ed. Net wt. 1 SDK 100% real inference Nutrition Facts Serving size: 1 Sipp Client Amount per pour Dependencies 0 Cold start < 1000ms Engine Rust · C++ · GGML Backend Browser-native WebGPU % Dev Value Open source 100% Type-safe 100%...

**来源信息**
- **来源**：Hacker News：AI 热帖
- **分类**：tip
- **发布时间**：2026-06-24 22:52（北京时间）
- **原文**：[打开原文](https://www.sipp.sh/)