Welcome to Inference Providers on the Hub 🔥

Hugging Face Blog · 1 year ago

Hugging Face integrated four serverless inference providers—fal, Replicate, SambaNova, and Together AI—directly into its Hub platform and client SDKs to offer unified access to model inference. PRO users receive $2 monthly in inference credits that work across all providers, with no markup added to standard provider API rates. Users can now run models like DeepSeek-R1 through their preferred provider using either their own API keys or Hugging Face's routing service, with configuration available through the website, Python, JavaScript, or HTTP calls.