vLLM

vLLM is a high-throughput and memory-efficient inference and serving engine for large language models (LLMs).

$100,000.0 - Collective