Reached #30 on 2026-05-05
github repository
youssofal/MTPLX
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.
Repository facts collected
- Current stars
- 2,720
- Stars in latest 7-day window
- 28
- Stars in latest 30-day window
- 125
Recent repository activity
youssofal/MTPLX added 28 stars from 2026-11-12 to 2026-11-18 (UTC).
The latest 7-day window recorded 3 more star additions compared with 2026-11-05 to 2026-11-11 (UTC). Star additions measure attention, not software quality or adoption.
Weekly rank: #157 in BeWisp’s tracked GitHub cohort on 2026-11-18 (UTC).
Latest recorded push: 2026-11-07 (UTC).
Latest recorded release: v2.12.3, published 2026-11-06 (UTC).
Star history
youssofal/MTPLX added 125 stars from 2026-10-20 to 2026-11-18 (UTC).
Source: GitHub star history. Latest collection: .
Periods end on the latest loaded date. Missing dates remain gaps. Use the slider or arrow keys to inspect a day.
Rank history
Daily, weekly and monthly positions in BeWisp’s tracked GitHub cohort.
2026-10-20 to 2026-11-18 (UTC). daily: #103; weekly: #157; monthly: #166 on 2026-11-18.
Achievements
Best positions reached in the preserved activity record.
Reached #63 on 2026-05-10
Reached #87 on 2026-05-29
Repository activity
Development signals collected separately from star activity.
Repository facts
- Language
- Python
- Licence
- Apache-2.0
- Forks
- 213
- Created
- 2026-05-02
Owner
youssofal · user.
Explore related repositories
These tracked repositories share topics or a primary language. Shared metadata does not establish that projects are interchangeable.
- jundot/omlx
LLM inference server with continuous batching & SSD caching for Apple Silicon — managed from the macOS menu bar
- decolua/9router
Unlimited FREE AI coding. Connect Claude Code, Codex, Cursor, Cline, Copilot, Antigravity to FREE Claude/GPT/Gemini via 40+ providers. Auto-fallback, RTK -40% tokens, never hit limits.
- iFurySt/open-computer-use
👾 Open Computer Use – Open-Source Alternative to Codex Computer Use