Cerebras Systems has partnered with Hugging Face to enhance AI inference capabilities, providing developers access to high-speed AI services. This collaboration aims to deliver an instant 10x speedup for applications involving AI chat, reasoning, and agentic apps. Cerebras Inference operates at over 2,000 tokens per second, which is 70 times faster than leading GPU solutions. The partnership will make popular models, including Llama 3.3 70B, available to Hugging Face developers, facilitating seamless API access to the Cerebras CS-3 powered AI infrastructure. Additionally, Cerebras is expanding its datacenter to support this growth in AI inference services.