G

Groq

Groq runs open-source LLMs (Llama, Mixtral) at record speeds on their custom LPU hardware, offering the fastest commercial LLM inference API.