Create API Endpoint from hugging face space

Question

I have created a Space on Hugging Face that runs my custom machine-learning model using Gradio. It works perfectly in the web interface, but now I want to convert this Space into an API endpoint that I can call from my application.

Could someone guide me through the process of converting my Hugging Face Space to an API endpoint? Specifically, I am looking for:

Steps to set up an Inference Endpoint from my existing Space. How to handle authentication and API keys for accessing the endpoint? Any best practices or tips for optimizing the performance of the endpoint.

Kevin · Accepted Answer · 2025-09-30 04:25:52Z

-1

Hugging Face Gradio Spaces has a built-in RESTful API.

On the bottom of your Space, there will be a link "Use via API" which you can click on that will provide instructions to access the space via API.

Alternatively you can access the API via HTTP requests (replacing the content within {{ }} with your actual HF info):

POST: https://{{your-hf-account-name}}-{{your-space-name}}.hf.space/gradio_api/call/predict

header: {Content-Type": "application/json"}

body:

{
  "data": ["{{your-input-prompt}}"]
}

Which will return an event ID

Then plug the event ID into:

GET: https://{{your-hf-account-name}}-{{your-space-name}}.hf.space/gradio_api/call/predict/dbb5d1d58e6b40f09ad07788f89efa6a{{EVENT-ID}}

header: {Content-Type": "application/json"}

Which will return the output.

answered Sep 30 at 4:25

Kevin

295 bronze badges

Sign up to request clarification or add additional context in comments.

Collectives™ on Stack Overflow

Create API Endpoint from hugging face space

1 Answer 1

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

Comments

Your Answer

Sign up or log in

Post as a guest

Related