MTPLX
youssofal
The fastest way to run Qwen 3.8 Flash Next, Qwen 3.8 27B and Ternary Bonsai 2 27B on a Mac: 125 tok/s in OpenCode on an M5 Max, and a 27B model on 16 GB Macs. Native MTP speculative decoding on Apple Silicon, exact at any temperature. OpenAI and Anthropic compatible local server.
Growth over time
2 data points · 2026-10-02 → 2026-10-25How do you feel about this project?
Ask AI about MTPLX
Question copied to clipboardWhat is the youssofal/MTPLX GitHub project? Description: "The fastest way to run Qwen 3.8 Flash Next, Qwen 3.8 27B and Ternary Bonsai 2 27B on a Mac: 125 tok/s in OpenCode on an M5 Max, and a 27B model on 16 GB Macs. Native MTP speculative decoding on Apple Silicon, exact at any temperature. OpenAI and Anthropic compatible local server.". Written in Python. Explain what it does, its main use cases, key features, and who would benefit from using it.
Question is copied to clipboard — paste it after the AI opens.
🏆 Show off this rank
Add it to a GitHub README or profile — the card updates as the ranking changes:
How to clone MTPLX
Clone via HTTPS
Clone via SSH
Download ZIP
Download main.zipFound an issue?
Report bugs or request features on the MTPLX issue tracker:
Open GitHub Issues