vLLM
A high-throughput and memory-efficient inference and serving engine for LLMs
vllm-project
Cost-efficient and pluggable Infrastructure components for GenAI inference
