Hugging Face Launches Inference Providers Hub Aggregating 10 External APIs

Hugging Face introduced Inference Providers, a unified hub feature that lets developers call models hosted on Together AI, Fireworks, Replicate, Cerebras, and six other inference providers through a single standardized API key and SDK. Switching providers requires changing one parameter in the InferenceClient, enabling cost comparison and fallback routing. Over 10,000 models are immediately supported across providers at launch.