Groq, a startup known for its blazing-fast AI inference hardware, has announced its availability as a provider on Hugging Face's new Inference Providers platform. This integration gives developers direct access to Groq's Language Processing Units (LPUs) from within the Hugging Face ecosystem.
"Groq's arrival on Hugging Face marks a significant step in making high-performance inference accessible to AI developers worldwide," said a company spokesperson.
Unlike traditional GPUs, Groq's LPUs are designed to process large language models with minimal latency. The partnership allows users to try models via Hugging Face's playground interface or through programmatic API calls. Models currently supported include Llama 3, Mixtral, and Gemma, with more expected to be added.
This move comes as part of Hugging Face's broader strategy to create a marketplace for inference providers, competing with cloud giants like AWS and Azure. For Groq, it provides a direct channel to reach millions of developers on the Hugging Face platform.