Pricing Plan of Microsoft Custom Speech Recognition

Question

I have been looking into Microsoft Custom Speech Recognition(CRIS)and I had a query with regards to its pricing plan. On their pricing plan, Microsoft has mentioned the following details for S0 plan:

Speech Translation - $2.50 per hour Speech to Text - $1 per hour Speech to Text with Custom Speech Model - $1.40 per hour Speech Endpoint Hosting - $40.00/model/month

My question may be amateurish since I have just begun to look into this but when they mention per hour, does this count the hours even when I am not hitting my speech endpoint, i.e does the pricing start as and when I create my resource?

For example - If I have created my Speech to Text service and have only hit it once in 24 hours, will I be charged only for the time it took to convert my speech to text or will I be charged 24 * $1 regardless of the usage?

Answer 1

You only get charged for what you use, not the up-time of the service.

Answer 2

the real-time of you audio is measured. If you want to transcribe 20 minutes of audio, this is what you are getting charged for. Be aware - if you try to transcribe 20 minutes of silence, it is a 20 minute charge...