Megadose The AI curve, compressed.

Hugging Face moves Inference Providers routing to OpenRouter

· TestingCatalog ·
Failed language model requests now retry with another provider inside the same call, but custom provider API keys no longer work for language models.

Developers keep their existing Hugging Face token and API, while OpenRouter selects providers based on live price, speed and reliability. Routing can preserve prompt caches, and users can still choose a specific provider or favor speed or price. Pricing stays at the provider’s rate with no markup on the Hugging Face bill. Image, video and audio generation continue to run directly through Hugging Face.

TestingCatalog's note

score 5

Categories: Money & Moves