Hugging Face Launches Inference Providers Marketplace with Pay-Per-Token Billing Across 10 Vendors

Hugging Face released an Inference Providers marketplace that aggregates access to over 10 compute providers—including Together AI, Replicate, and Fireworks AI—under a single API key and unified pay-per-token billing. Developers can switch providers per request for cost or latency optimization without changing integration code. Over 500 open models are available at launch through the unified endpoint.