# Pricing for Huggingface Endpoint **URL:** \ **Category:** Inference Endpoints on the Hub\ **Created:** [September 4, 2023, 9:18am UTC](https://discuss.huggingface.co/t/pricing-for-huggingface-endpoint/53456 "2023-09-04T09:18:10Z")\ **Posts on this page:** 7\ **Page:** 1
**Author:** ![baniasbaabe](https://avatars.discourse-cdn.com/v4/letter/b/76d3ee/32.png) [@baniasbaabe](https://discuss.huggingface.co/u/baniasbaabe)\ **Post date:** [September 4, 2023, 9:18am UTC](https://discuss.huggingface.co/t/pricing-for-huggingface-endpoint/53456/1 "2023-09-04T09:18:10Z")
I am a bit confused about the pricing model. Let’s say I deploy a model on a CPU Basic machine ($0.06/hour). So do I pay as long as the model is deployed or do I pay only for the compute time (e.g. I make 2 requests and every request takes 10 seconds to run, so do I only pay for the 20 seconds). I ask because I want to do a hobby project but don’t want to pay a lot for it. Alternatives like cerebrium only offer 3 models for their starter version :') ---
**Author:** ![philschmid](https://sea2.discourse-cdn.com/hellohellohello/user_avatar/discuss.huggingface.co/philschmid/32/38022_2.png) [@philschmid](https://discuss.huggingface.co/u/philschmid)\ **Post date:** [September 4, 2023, 11:12am UTC](https://discuss.huggingface.co/t/pricing-for-huggingface-endpoint/53456/2 "2023-09-04T11:12:27Z")
You can find the information in the documentation: [Pricing](https://huggingface.co/docs/inference-endpoints/pricing) ---
**Author:** ![had123](https://avatars.discourse-cdn.com/v4/letter/h/ecccb3/32.png) [@had123](https://discuss.huggingface.co/u/had123)\ **Post date:** [November 21, 2023, 11:20am UTC](https://discuss.huggingface.co/t/pricing-for-huggingface-endpoint/53456/3 "2023-11-21T11:20:23Z")
Hello, I have the same question as @baniasbaabe, I don’t understand how you calculate the price. Can you please give us more clarity? If I subscribe for one month, will I have to pay for 730 hours or only the number of hours that I have interactions with the endpoint? ---
**Author:** ![YKinebas](https://sea2.discourse-cdn.com/hellohellohello/user_avatar/discuss.huggingface.co/ykinebas/32/22353_2.png) [@YKinebas](https://discuss.huggingface.co/u/YKinebas)\ **Post date:** [December 21, 2023, 8:24pm UTC](https://discuss.huggingface.co/t/pricing-for-huggingface-endpoint/53456/4 "2023-12-21T20:24:30Z")
From the link above: At the end of the subscription period, the user or organization account will be charged for the compute resources used while Endpoints are _initializing_ and in a _running_ state. So only compute time ---
**Author:** ![DigitalMantis](https://sea2.discourse-cdn.com/hellohellohello/user_avatar/discuss.huggingface.co/digitalmantis/32/28072_2.png) [@DigitalMantis](https://discuss.huggingface.co/u/DigitalMantis)\ **Post date:** [March 10, 2024, 5:35pm UTC](https://discuss.huggingface.co/t/pricing-for-huggingface-endpoint/53456/5 "2024-03-10T17:35:11Z")
so does it cost less to keep a endpoint in a running state? ---
**Author:** ![meganariley](https://sea2.discourse-cdn.com/hellohellohello/user_avatar/discuss.huggingface.co/meganariley/32/56966_2.png) [@meganariley](https://discuss.huggingface.co/u/meganariley)\ **Post date:** [March 22, 2024, 9:21pm UTC](https://discuss.huggingface.co/t/pricing-for-huggingface-endpoint/53456/6 "2024-03-22T21:21:01Z")
Hi @DigitalMantis In order to stop incurring cost on running endpoints, you’ll need to delete or [pause](https://huggingface.co/docs/inference-endpoints/guides/pause_endpoint) them. We also have an [automatic scaling to 0](https://huggingface.co/docs/inference-endpoints/autoscaling#scaling-to-0) feature! You can check usage and billing at any time in your [billing settings](https://huggingface.co/settings/billing). Hope this helps! ---
**Author:** ![langhoangal](https://sea2.discourse-cdn.com/hellohellohello/user_avatar/discuss.huggingface.co/langhoangal/32/40404_2.png) [@langhoangal](https://discuss.huggingface.co/u/langhoangal)\ **Post date:** [February 5, 2025, 12:34am UTC](https://discuss.huggingface.co/t/pricing-for-huggingface-endpoint/53456/7 "2025-02-05T00:34:04Z")
After deployment, I tested with 9 request within few minutes. And later (after few hours) I saw the compute time is 46 minutes. Trying another service for now.