> ## Documentation Index
> Fetch the complete documentation index at: https://docs.weflayr.com/llms.txt
> Use this file to discover all available pages before exploring further.
# AI Cost
> What each feature of your product costs, per run and per customer.
A feature is the `feature_name` your calls carry ([gateway](/api/gateway/responses#weflayr-fields) or [SDK](/sdks/python#propagate_metadata)). A feature run is one execution of that feature, whatever the number of LLM calls it makes.
## Filter and compare
Filter by period, model, and by any [customer tag](/product/customer-tags) or tag your calls carry. The chart shows the total cost or the average cost per feature run, by day, week, month, quarter or year, and downloads as a CSV.
The table lists every feature with its cost, share of spend, number of runs, tokens, average cost per run and month-over-month change. Expand a row to see its calls.
## Feature detail
Click a feature to see its cost and adoption over time, its top customers and the models it runs on.
This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.