MTPLX
Run local LLMs on Apple Silicon, around twice as fast.
https://github.com/youssofal/MTPLX
GitHub
GitHub - youssofal/MTPLX: 3x faster speeds on MLX | Qwen 3.8 27B | Native MTP Speculative Decoding On Apple Silicon With No External…
3x faster speeds on MLX | Qwen 3.8 27B | Native MTP Speculative Decoding On Apple Silicon With No External Drafter. - youssofal/MTPLX
t.me/PythonHub/50939Translating to English…
August 27, 2026 42 1
