A WireMock simulation of Hugging Face's REST API — covering the curated model catalog with its accelerator and engine recipes, filtered listings, and one-click deploys of catalog models or recipes.
A shared instance of this template runs on a public endpoint — call it right now from curl, an SDK, or an AI agent. Responses are fixed examples and behavior can't be changed; rate limits apply.
Jump to the endpoints ↓Run this template as a mock API in WireMock Cloud for a private instance you control — edit stubs, add stateful scenarios, use dynamic response templating, and inspect requests in the log.
Open in WireMock Cloud (free) →Live sandbox · no signup, no API key
A running WireMock Cloud instance of this template, callable right now from a terminal, a script, or an AI agent.
Endpoints
GET List catalog items [listCatalogItems]
curl -s -X GET 'https://k7joz5.wiremockapi.cloud/api/v1/catalog/list'Spec catalog — Hugging Face Endpoints API (1.0.0) · 3 endpoints
| Method | Path | Summary |
|---|---|---|
| GET | /api/v1/catalog/list | List catalog items |
| POST | /api/v1/catalog/model/{author}/{model}/deploy | Deploy a catalog model as an Inference Endpoint |
| POST | /api/v1/catalog/recipe/{recipePublicId}/deploy | Deploy a specific catalog recipe as an Inference Endpoint |
Unauthenticated WireMock Cloud demo sandbox for Hugging Face — not an official Hugging Face sandbox, and it returns simulated example data only.
The template serves the catalog API at endpoints.huggingface.co, the list of curated models Inference Endpoints can deploy in one click. Listings replay the live catalog as recorded, model order and hidden filter behavior included. A deploy builds the endpoint the catalog would: a GPU deploy of Qwen3-1.7B comes back on an nvidia-l4 instance with a vLLM container and your chosen name. Tools that browse the catalog and launch endpoints from it can be exercised without starting anything billable.
Because the live service filters on attributes the response doesn't show, and the template keeps that behavior. A GPU filter leaves out some models that list a GPU recipe and includes a few that list none. License filters take ids such as apache-2.0, not the displayed names, so license=MIT returns nothing.
Case-insensitively, ignoring hyphens, underscores, slashes and spaces, so a search for QWEN3 also surfaces distilled models with Qwen3 in their repo id.
You need an Authorization header, because the live API checks for one before anything else, and so does the mock. Any bearer value passes. Never send a real hf_ token to the shared sandbox.
Not through this API, which has no route to read deployments back. Each deploy is recorded in the mock's state, and endpoint management lives in the Hugging Face Inference Endpoints template.
Vendor names identify APIs represented by WireMock template sources. This page does not imply vendor endorsement, certification, partnership, or official integration status.