youssofal/MTPLX star history
PythonThe fastest way to run Qwen 3.8 Flash Next and Qwen 3.8 27B on a Mac: 125 tok/s in OpenCode on an M5 Max. Native MTP speculative decoding on Apple Silicon, exact at any temperature. OpenAI and Anthropic compatible local server.
Why it’s trending
youssofal/MTPLX is accelerating — star growth is speeding up by +10.0/day (95th percentile in the tracked cohort), on top of +10.0 stars/day. A recent release is helping fuel the run.
Star acceleration +10.0/day (95th pctl)Star velocity +10.0/day (69th pctl)Fork velocity +1.7/day (72th pctl)New release published
Figures drawn only from Breakwave data
Breakout forecast14-day horizon
87%breakout probability
Confidence11%
Momentum85~Momentum Score (0–100): how fast this repo is accelerating compared with every other tracked repo — not how big it is. Provisional: based on limited history so far, so treat it as an early read.provisional
Why it ranks here
Star acceleration95pctl
Star velocity69pctl
Fork velocity72pctl
Contributor growth100pctl
Stars
2.4k
Forks
189
Watchers
20
Open issues
30
★ / day
+10.0
trailing 7d
Acceleration
+10.0
▲ velocity
Star growth over time
Reconstructed from 28 snapshots — the time-series GitHub’s API doesn’t expose.
Latest events
- New release publishedSep 24, 2026
Release count rose to 54.
Add this to your README
Live score, refreshed every 6 hours. Links back to this page.
[](https://breakwave.vercel.app/repo/youssofal/MTPLX)