**DeepSeek V4 Flash, up to 32 tok/s on AMD Ryzen AI MAX+ 395**

> AMD-Powered Lucebox runs the 284B DeepSeek V4 Flash model locally on AMD Ryzen AI MAX+ 395 with 128 GB unified memory, reaching up to 32 tok/s decode and roughly 250 tok/s sparse prefill. July 2026 By Davide Ciffa DeepSeek V4 Flash: 284B model, up to 32 tok/s on AMD Ryzen AI MAX+ 395 AMD-Powered Lucebox runs the full DeepSeek V4 Flash target locally on AMD Ryzen AI MAX+ 395 with 128 GB unified memory: up to 32.0 tok/...

**来源信息**
- **来源**：Hacker News：AI 热帖
- **分类**：ai-products
- **发布时间**：2026-07-28 23:03（北京时间）
- **原文**：[打开原文](https://www.lucebox.com/blog/deepseek-v4-strix-halo)