Local LLM Speed Lab

Decode throughput of local AI models on home hardware