A programmable Mixture-of-Models router for heterogeneous LLM inference
vllm-project/semantic-router is drawing steady momentum at +8.7 stars/day (66th percentile in the tracked cohort), for a momentum score of 45.7/100. Forks are climbing too, at +3.9/day.
Low breakout odds over the next 14 days, led by push recency (100th pctl) and fork velocity (84th pctl). Confidence is high given the available history.
Reconstructed from 205 snapshots — the time-series GitHub’s API doesn’t expose.
~13 forks/day, 3.5× the trailing baseline.
~12 forks/day, 3.4× the trailing baseline.
~14 forks/day, 3.6× the trailing baseline.
Live score, refreshed every 6 hours. Links back to this page.
[](https://breakwave.vercel.app/repo/vllm-project/semantic-router)Transparent heuristic · logged for model training
Percentile rank within the tracked cohort. The score self-calibrates — it answers “accelerating vs. everything else,” not raw size.
~10 forks/day, 3.8× the trailing baseline.
~12 forks/day, 4.0× the trailing baseline.
~17 forks/day, 3.0× the trailing baseline.
~14 forks/day, 7.8× the trailing baseline.
~13 forks/day, 10.6× the trailing baseline.