Hugging Face Inference

Open models from Hugging Face's inference providers through one router: chat, embeddings, images, speech.

API keyCopy Anchor Link

1. Open huggingface.co and create an API key with only the access your apps need.
2. Copy it. It may be shown only once.

What to create: A fine-grained access token from Hugging Face settings with the Inference Providers permission.

Add it to your Gateway. It asks for the key:

everyapp provider add huggingface

Use it in an appCopy Anchor Link

In the app's everyapp.config.ts:

capabilities: {
  huggingface: { access: "read", why: "What the app does with it." },
},

The app calls Hugging Face Inference's API with integrationFetch(), and the Gateway adds the key. More in API providers.

Check it worked: everyapp provider show huggingface shows the key as set.

Cost: Pay as you go at each provider's own rate with no markup; PRO accounts get $2 of compute credits a month.

Something wrong or missing? Open an issue.