Medusa
Medusa: Simple Framework for Accelerating LLM Generation with Multiple Decoding Heads
FasterDecoding/Medusa
View on GitHub ↗Medusa: Simple Framework for Accelerating LLM Generation with Multiple Decoding Heads
Medusa grew +0.1% in the last 30 days, ranks 66th of 490 in Local Runtime / Inference Engine, last commit 849 days ago.
Star trajectory
123 days tracked2026-06-212.8k stars today2026-10-23
Momentum
+0.1% in 30d, below the Local Runtime / Inference Engine median of +1.7%.
Maintenance
Last commit 849 days ago, staler than the median of 5.5d.
How it compares in Local Runtime / Inference Engine
Medusa
2.8k
stars
+0.1% 30d
commit 849 days ago
Local Runtime / Inference Engine median
414.5
stars
+1.7% 30d
commit ~5.5d ago
Leader · ollama
184.4k
stars
+1.6% 30d
commit today
Ranked 66th of 490 by stars.
dmtrKovalenko/fframes2.9k-148deeplethe/forkd2.9k-126
FasterDecoding/Medusa2.8k
lightseekorg/tokenspeed2.8k+6janhq/nitro2.7k+27- Language
- Jupyter Notebook
- License
- Apache License 2.0
- Contributors
- 11
- Forks
- 206
- Watchers
- 29
- Open issues
- 51
- Releases
- 1
- Created
- 2023-09-10
- Stars this week
- +0
You've seen where Medusa stands today.
Get the moves before everyone else: the 5 Local Runtime / Inference Engine repos climbing fastest, every Tuesday.
Free weekly digest. Double opt-in, unsubscribe anytime.