Billing

Usage

Usage is calculated per-request, at the advertised cost when the request is finalised. The cost of each request is calculated based on the total amount of input tokens and output tokens that request clocked in at. As requests are made, aggregate usage and cost, broken down by API key and separately by model, will land in the SiteHost Control Panel:

Aggregate metrics in the SiteHost Control Panel can be delayed by a number of minutes.

Some models (such as embedding models) do not generate output tokens as they always create fixed vectors. For these models, you will always see 0 output token usage and your cost will be based entirely upon input token usage.

Breakdowns

We store and expose the full request breakdown on a per-key, per-hour basis for three months, after which the total is the only remaining record. If you'd like to see the full breakdown for a given key and hour, you can view it in the SiteHost Control Panel under the usage tab when viewing an API key, then selecting the hour in question.

Invoicing

At the start of each month, an invoice will be created for your AI usage from the previous month, as a total of all usage rounded to the nearest cent. To view a breakdown of your usage, please visit the SiteHost Control Panel and select the month in question, which will give you a per-key and a per-model breakdown.