vllm-project/vllm

A high-throughput and memory-efficient inference and serving engine for LLMs

72,395 stars14,085 forksPythonView on GitHub ↗
7-day Growth RateA+
+32/day
Top 3.6%

Daily compound star growth rate over 7 days

30-day Growth RateA+
+1/day
Top 3.5%

Daily compound star growth rate over 30 days

AccelerationS
Steady
Top 0.0%

Is the growth rate speeding up or slowing down?

OriginalityB
3/100
Top 34.5%

Stars earned relative to the number of similar repos in the same category

Topics

amdblackwellcudadeepseekdeepseek-v3gptgpt-ossinferencekimillamallmllm-servingmodel-servingmoeopenaipytorchqwenqwen3tputransformer

Data as of 2026-03-08