The rule, in one sentence
One call = ceiling( duration_in_seconds / 60 ) minutes. A 1-second call is 1 minute. A 60-second call is 1 minute. A 61-second call is 2 minutes.
Why round up?
It's how every telecom carrier in the U.S. bills us, so we have to pass it through. We chose not to invent fractional-minute billing because:
- It would make the invoice impossible to reconcile against the carrier bill.
- The math doesn't actually save customers much — most healthcare calls run 90 seconds to 4 minutes, where rounding is in the noise.
- A predictable "minutes used" number is easier to plan against than "12.387 minute-equivalents."
What is and isn't billed
| Event | Billed? | Notes |
|---|---|---|
| AI answers and talks to a caller | ✅ | Rounded up to the next whole minute. |
| Caller hangs up before AI speaks (≤ 2s) | ❌ | Not billed. |
| Caller hangs up while AI is greeting (3–10s) | ✅ | Counts as 1 minute. |
| Voicemail with no message left | ❌ | Not billed. |
| Voicemail with a message left | ✅ | Message duration, rounded up. |
| Call transferred to a human | ✅ (AI portion only) | The forwarded leg is billed by your carrier. |
| Outbound call placed by your team via the web dialer | ✅ | Same per-minute rate. |
| Outbound AI follow-up call (e.g., reminder confirmations) | ✅ | Counts as a minute even if it goes to voicemail and leaves a message. |
| Test calls you make to your own number | ✅ | There is no separate test mode. |
| Internal staff-to-staff calls inside FrontDesk | ❌ | Not metered. |
Inbound vs outbound
On Starter and Professional plans, inbound and outbound AI minutes share one monthly pool. Enterprise plans can split them on request — useful if your outbound campaigns are predictable and you want a separate cap.
SMS is billed separately, in segments — not minutes. See Understand usage and overages for the SMS rules.
Where to watch your usage
- Live usage: Settings → Billing → Usage. Shows minutes used, days remaining, and projected overage if you keep the current pace.
- Per-day breakdown: Same page — chart at the bottom. Useful for spotting a Monday spike or a robocall episode.
- Overage alerts: Settings → Billing → Usage caps. Set an email at 80%, 100%, or a custom threshold. We strongly recommend setting these rather than running blind.
- Hard cap (rarely recommended): Same page — toggle "Stop service at limit." This will drop real patient calls once you hit the cap, so most practices prefer a high soft cap with alerts instead.
What to do if you're trending over
- Look at the per-day breakdown — is it a steady increase (you're growing) or a one-day spike (a robocall, a campaign, a glitch)?
- If steady growth, upgrade one tier — the price difference is usually less than the overage rate.
- If a spike, check Calls → Logs for the day and filter by long calls. Robocalls and accidental loops show up immediately.
See Understand usage and overages for overage pricing and cap settings.