Groq

Groq's API provides ultra-fast inference for open-source language models like Llama 3, Mixtral, and Gemma using custom LPU hardware. Optimized for low-latency applications.