How Huggingface pricing works for model Deployment?

Inference Endpoints on the Hub
1

i want to know how pricing in hugging face works .i want to host a stable diffusion model on GPU server which is charging 0.6$ per hour …what does it mean ? if no one will use my App for image generation would i ll be billed or not .if i ll get billed how much i need to pay .? its very confusing …please someone help me out >

3 Likes
2

Please check out the documentation: Pricing There are also examples included.

3

This pricing link does not answer the question I also have. After further searching and reading the pricing doc, I think that I got an answer for you @nicky007 :
unfortunately you are charged when your app does not make requests

since every instance you create is ON (running) all the time. This makes using hugging face very expensive. Particularly if you have multiple models that you need to run.

2 Likes