Hugging Face's Inference Providers has a new default routing partner for language models: OpenRouter.
Hugging Face is consolidating its language-model inference through OpenRouter's API, moving from its own orchestration layer to OpenRouter's router. The change was confirmed by Hugging Face on November 16, 2026.
What changed
In the Inference Providers dashboard, developers can now search the catalog of language models served by OpenRouter. A new "OpenRouter models" section lists around 330 models and their capabilities, with the models ranked by popularity.
Inference Providers is Hugging Face's collection of serverless APIs that lets developers use models from different providers through a single interface.…