Inference Providers is Hugging Face's serverless inference layer. One Hugging Face token reaches open models for text, image, video and audio across partners such as Cerebras, Groq, Together AI, Replicate and fal, at the provider's price with no markup. Each request is routed to a provider serving the model and switches to another if one is down or rate limited.
It is a drop-in replacement for the OpenAI chat completions API (https://router.huggingface.co/v1) and also has huggingface_hub and JavaScript clients that cover other tasks such as text-to-image, embeddings and speech. Append :fastest or :cheapest to a model id to choose a routing policy, or name a provider.
It is pay-as-you-go, and PRO, Team and Enterprise plans include monthly credits.