Skip to content
August 2026

Unlimited AI Receptionist Minutes: Where the Break-Even Is

Every flat plan has a call volume at which the vendor starts losing money, and the component cost model computes it.

An unlimited plan is not a generous plan. It is a priced bet, and the vendor has run the arithmetic even if the pricing page does not show you any of it.

The bet is that your average usage sits far enough below the break-even that the heavy users are paid for by everybody else. That is not a criticism. It is how flat pricing works everywhere, from gym memberships to data plans, and it is only a problem when the buyer cannot see which side of the line they are on.

Turning the claim into a number

This site computes a floor of $0.0465 a minute from four dated published component rates. Divide any monthly price by it and you get the volume at which the vendor's revenue equals its variable cost.

Monthly priceVendor break-evenIn 4-minute callsPer working day
$491,054 min~264 calls~12 calls
$791,699 min~425 calls~19 calls
$1994,280 min~1,070 calls~49 calls
$3998,581 min~2,145 calls~98 calls
$4499,656 min~2,414 calls~110 calls
$598.8012,877 min~3,219 calls~146 calls

The right-hand column is the useful one. A $49 plan stops being profitable for the vendor at about twelve calls a day, and a $399 plan at about ninety-eight.

What that tells a buyer

Two things, and they point in opposite directions.

If you are well below the line, you are subsidising somebody else. A business taking 300 calls a month on a $399 plan is using about $56 of components and paying $399, which is fine if the integration and the support are worth the difference, and poor value if they are not. Metered pricing would serve you better and is available from the platform vendors.

If you are near or above the line, read the fair-use clause. This is the practical warning. A plan that is genuinely unlimited at ninety-eight calls a day is a plan the vendor loses money on, and vendors do not lose money quietly. What they do instead is one of three things.

  • A fair-use ceiling in the terms, which converts unlimited into a cap you find out about by exceeding it.
  • A concurrency limit, which caps calls in parallel rather than calls in total, and bites hardest exactly when your phone is busiest.
  • A migration conversation, where you are moved to a custom plan at renewal on the grounds that your usage is atypical.

The second is the one to ask about first, because it is a real engineering limit rather than a contractual one and it is almost never on the pricing page. On voice platforms elsewhere in this directory, concurrency is metered explicitly and separately from minutes: one vendor bills burst concurrency at double rate and then rejects the call, and another sells a five-call cap whose removal costs $100 a year at an identical per-minute rate.

The same argument from the seller's side

This site has already published the reseller version of this arithmetic, and it lands in the same place from the opposite direction. On a $297 flat retainer, a client using 500 minutes a month has a variable cost of roughly $42 to $48.

Which means about 85% of the gross profit is the retainer itself rather than any markup on minutes. Flat-rate voice is not a minutes business at typical volumes. It is a subscription business with a small variable cost attached, and understanding that is what lets you ask the right question at renewal.

The $0.0465 floor is computed from four dated published component rates and is not measured from any invoice; a vendor's true variable cost will differ with its own committed rates and will usually be lower. Break-even columns are arithmetic on that floor at the published prices in this lane. This page does not assert that any named vendor advertises unlimited minutes - that is not a structured field in this directory yet, and the arithmetic here applies to any flat monthly price whatever it is called.

Questions

Is unlimited AI receptionist pricing really unlimited?
It is unlimited up to a volume the vendor has calculated and you have not been shown. Against a $0.0465 component floor, a $49 plan reaches the vendor's break-even at about twelve calls a day and a $399 plan at about ninety-eight. Past that the limit arrives as fair use, concurrency, or a renewal conversation.
What is a concurrency limit and why does it matter more than minutes?
It caps how many calls can happen at the same time rather than how many in total, so it bites exactly when your phone is busiest. It is rarely on a pricing page. Elsewhere in this directory one platform bills burst concurrency at double rate then rejects the call, and another charges $100 a year to lift a five-call cap.
Is a flat plan or per-minute pricing better for a small business?
Per-minute is usually cheaper below the break-even, which most single-location businesses are. At 300 four-minute calls a month you use about $56 of components; a $399 flat plan is buying integration and support, not cheap minutes. Flat wins when volume is high or genuinely unpredictable.
How do I know if I am overpaying on a flat plan?
Divide your monthly price by $0.0465 to get the break-even minutes, then compare it to your actual usage. If you are using under a fifth of it, you are buying software and support rather than calls, and should price that bundle on its own merits against a metered alternative.
Why do vendors offer unlimited plans at all?
Because at typical volumes this is a subscription business with a small variable cost attached, not a minutes business. On a $297 retainer with 500 minutes used, roughly 85% of gross profit is the retainer itself. Unlimited is cheap to promise when most customers stay well under.

Tools mentioned

All tools

Sources

Source interests are labelled. Almost everything published about this subject is written by someone selling into it.

More from the blog

All posts