Qwen3.6-27B-MTPLX-Optimized-Speed · MTPLX · MacBook M5 Max
English | 简体中文 | 日本語 · ← All recipes
Source level: 📣 Claimed · Date: 2026-05-03 · Numbers come from the source. This repo did not measure them again.
Hardware
- MacBook M5 Max (Mac)
- Reference cost: Not published. See the source.
Model
- Qwen3.6-27B-MTPLX-Optimized-Speed ·
MLXCommunity 4-bit backbone + 3-bit draft LM head (MTP depth 3) - Download: huggingface.co/Youssofal/Qwen3.6-27B-MTPLX-Optimized-Speed
Engine
- MTPLX
- Note: MTPLX v0.1.0-preview
Launch command
brew install youssofal/mtplx/mtplx && mtplx start
Key parameters
- Context: —
- Concurrency: single request
- Tags:
3bit,4bit,dense,speculative,MTP
Expected speed
- 63.056 tok/s decode (📣 Claimed)
Caveats
- Not a first-hand measurement (claimed or relayed). Your results can differ.
- Speed changes with drivers, engine version, context length, and prompt length.
Source & attribution
- https://huggingface.co/Youssofal/Qwen3.6-27B-MTPLX-Optimized-Speed
- Credit to the original author. This page only summarizes and links. Data under CC BY 4.0.
Source file: recipes/hf-youssofal-qwen3-6-27b-mtplx-optimized-speed-1.md