What Does an Answering Service Cost? (2026 Breakdown)

If you've ever called an answering service to ask for a quote, you know the answer is almost never a straight number. It's "well, it depends on your volume," followed by a tier chart that needs its own decoder ring. So let's pull the bill apart and look at how the money actually moves.

This is the honest version. No invented prices, no scare tactics. Just the pricing models, some example math you can run against your own call count, and where the costs tend to sneak up on people.

The three ways answering services charge you

Almost every human-staffed service uses one of three billing models, or a blend of them.

• Per minute. You pay for the time agents spend on your calls, usually rounded up to the next minute or half minute. This is the most common model.
• Per call. You pay a flat amount each time a call comes in, no matter how long it runs.
• Per month with a minimum. You buy a bucket of minutes or calls up front, and you pay for the whole bucket whether you use it or not.

Most providers also stack a base fee on top, plus setup charges, and sometimes a per-account fee just to keep your line active. So the headline rate is rarely the real rate.

Per-minute pricing, explained

This is the one most people mean when they ask how much does an answering service cost. You're billed for connected agent time. The catch is what counts as "time." A lot of services bill in rounded increments, so a 40-second call can bill as a full minute, and they often include hold time, wrap-up notes, and the few seconds of dead air at the start.

Here's clearly labeled example math (not a real quote, just round numbers to show the shape):

Say a service charges $1.50 a minute and you get 300 calls in a month that average 2 minutes each. That's 600 minutes, or about $900 before any base fee. Add a $100 base fee and you're near $1,000 for one ordinary month.

Now picture a slow month with 150 calls. Your bill drops, but the base fee stays. Per-minute billing rewards quiet months and punishes loud ones, which matters more than the rate itself.

Per-call pricing, explained

Per-call billing trades the stopwatch for a flat ticket price. You pay the same whether the caller talks for 20 seconds or four minutes. That's predictable when your calls run long, and it can be a bad deal when most of your calls are quick hang-ups or wrong numbers, because you pay full price for a 15-second "sorry, wrong office."

Run the same example shape. At a made-up $2.50 per call with 300 calls, you're at $750 plus the base fee. Cheaper than the per-minute example above, but only because we assumed long calls. Flip the average call length and the math flips with it.

Monthly minimums, explained

This is where a lot of small businesses quietly overpay. You commit to a plan, say 500 minutes a month, and you pay for all 500 even if you use 200. Unused minutes usually don't roll over. The plan looks cheap per minute on paper, but if your real usage sits well under the bucket, your effective rate is much higher than the brochure says.

And if you go over the bucket? Overage minutes almost always cost more than your plan rate. So the minimum protects the provider on both ends.

Why busy months get expensive fast

Here's the part the tier chart doesn't advertise. Your call volume isn't flat. A plumber's phone explodes during a cold snap. A clinic's lines jam during flu season. A contractor gets buried after a storm. Those are exactly the months you most need every call answered, and they're exactly the months a per-minute or overage bill spikes.

The pricing models all share one trait: cost scales with volume. When you're slow, you pay a base fee for not much. When you're slammed, the meter runs hot right when cash is tight from other pressures. You don't get a volume discount for your worst week. You get an overage line.

That volatility is the real cost of an answering service, not the per-minute number. It's hard to budget for a bill that swings a few hundred dollars based on the weather.

Where flat-rate AI answering fits

An AI receptionist prices differently because the cost structure underneath it is different. There's no agent on a stopwatch, so there's no reason to bill by the minute. Most AI answering runs on a flat monthly plan: you pay the same in a quiet month and a brutal one.

That changes the math in your favor exactly when you'd be hurting under per-minute billing. A storm week that would've blown past your minute bucket just gets answered. The bill doesn't move.

There's a second cost most quotes ignore: what happens after the call. A human service takes a message and emails it to you. With OneBy, every call gets a written summary and turns into an assigned task or ticket, so the follow-up doesn't fall through the cracks. The "cost" of a missed callback (a job you never booked) usually dwarfs the per-minute rate, and that's the line item nobody puts on the quote.

We broke the AI side down in more detail in the real cost of an AI receptionist, including where AI is and isn't the right call. And if you're weighing a managed service against software, OneBy vs Smith.ai lays out how the two approaches actually differ.

A quick checklist before you sign anything

Whatever model you're quoted, ask these before you commit:

• What's the billing increment? Per second, per minute, or rounded up?
• Does hold time and wrap-up count as billable time?
• What's the base fee, the setup fee, and any per-account fee?
• Do unused minutes roll over, and what's the overage rate?
• What actually happens to the message after the call ends?

That last question is the one that separates a service that takes a note from one that moves the work forward.

So, what does an answering service cost?

The honest answer is that answering service pricing is less about the rate and more about the model and your volume. Per-minute is fair when you're quiet and rough when you're busy. Per-call is the opposite. Monthly minimums look cheap until you under-use them. And almost every model gets expensive in the exact months you can least afford it.

Flat-rate AI answering takes the volatility out. You trade a swinging bill for a steady one, and you get the call turned into something you can act on instead of a voicemail you'll forget. Run the example math against your own call count. The shape of your year usually tells you which model you're built for.

See how flat-rate answering would handle your busiest month. Book a 10-minute demo.