> ## Documentation Index
> Fetch the complete documentation index at: https://docs.recepta.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Usage & overages

> Tracking calls, conversations, and escalations against your plan.

The **Usage** page shows where you stand against your plan's allowances for the current billing period, so a large invoice is never a surprise.

## What's metered

| Metric                   | Counted as                                             | Overage                              |
| ------------------------ | ------------------------------------------------------ | ------------------------------------ |
| Calls                    | One per answered call, inbound or outbound, any length | $1.95–$2.75 per call by tier         |
| Chat & SMS conversations | One per conversation thread, not per message           | $0.45–$0.75 per conversation by tier |
| Human escalations        | One per transfer to a person                           | \$3 each, all tiers                  |

Exact rates per tier are on the [Plans](/plans) page.

## Reading your usage

<Columns cols={2}>
  <Card title="Mid-period pace" icon="gauge-high">
    Halfway through the period, you should be around half your allowance. Well ahead of that means you'll go over — worth knowing before the invoice.
  </Card>

  <Card title="Escalation rate" icon="phone-arrow-up-right">
    A rising share of calls transferring to a human usually means a knowledge gap, not a busier month. See [QA & learning](/receptionist/qa-and-learning).
  </Card>
</Columns>

## Overages aren't a penalty

Going slightly over costs less than jumping a tier early. Move up when your overages consistently exceed the gap between tiers — not the first month you exceed your allowance.

<Note>
  Overages are calculated and billed at the end of the billing period, itemized on your invoice by category. See [Billing](/account/billing).
</Note>

## Bringing usage down

If your call count is higher than expected, the cause is usually one of:

<AccordionGroup>
  <Accordion title="Outbound campaigns at volume" icon="bullhorn">
    Every outbound attempt is a call. A large campaign will move the number materially — budget for it before launching. See [Campaigns](/outbound/campaigns).
  </Accordion>

  <Accordion title="Repeat callers" icon="rotate">
    The same customer calling three times about one job is three calls. Often it means the first call didn't resolve — read those transcripts.
  </Accordion>

  <Accordion title="Spam and robocalls" icon="ban">
    These reach your number like anything else. If the volume is significant, tell [info@recepta.ai](mailto:info@recepta.ai).
  </Accordion>

  <Accordion title="Testing" icon="flask">
    Your own test calls count. Worth knowing during a heavy configuration week, though it's rarely material.
  </Accordion>
</AccordionGroup>

## Over the API

Usage is exposed via the API, so you can pull it into your own dashboards or alert when you cross a threshold. Combine with the `call.ended` [webhook](/api-reference/webhooks) if you want real-time counting.