---
format: "aidr-story-markdown/v1"
id: "2655af8eb0bc114b5bc4b3a5da09da4d508b4569e0bf29a7c1484a8026cfc7b9"
canonical_url: "https://aidr.today/2655af8e?lang=en"
title: "DFlash 2 Hits 70 Tok/s on M5 Max MacBook Pro in 4.6x Speed Jump"
lang: "en"
requested_lang: "en"
available_langs: ["en","vi"]
translation_fallback: null
fallback_fields: []
published_at: "2026-08-18T23:50:02.000Z"
category: "Infra"
topics: ["qwen","inference","apple","hardware"]
source_urls: ["https://huggingnews.com/ai/dflash-2-hits-70-toks-on-m5-max-macbook-pro-in-46x-speed-jump-01fefd72"]
summary: "Inco AI and Z Lab launched a new inference engine to accelerate the execution of large language models on Apple hardware. DFlash 2 allows the Qwen3.8-27B model…"
---

# DFlash 2 Hits 70 Tok/s on M5 Max MacBook Pro in 4\.6x Speed Jump

> [Open the canonical story](<https://aidr.today/2655af8e?lang=en>)

**Published:** 2026-08-18T23:50:02.000Z
**Category:** Infra
**Topics:** qwen, inference, apple, hardware

## Summary

Inco AI and Z Lab launched a new inference engine to accelerate the execution of large language models on Apple hardware\. DFlash 2 allows the Qwen3\.8\-27B model…

## Sources

- [Story source](<https://huggingnews.com/ai/dflash-2-hits-70-toks-on-m5-max-macbook-pro-in-46x-speed-jump-01fefd72>)