AI customer service pricing usually falls into one of three shapes: a per-seat fee, a per-resolution fee, or a flat platform fee with metered usage on top. Each one moves the risk of an unpredictable month onto a different party, and knowing which is which matters more than the headline number on the pricing page.

Why pricing in this category is hard to compare

Two vendors can advertise similar-looking numbers and mean completely different things by them. One charges per seat regardless of volume. Another charges per "resolution", a word with no fixed industry definition. A third charges a flat fee with usage bundled in. None of these are dishonest, exactly, but none of them are directly comparable without understanding what each one does to your risk in a bad month.

The three main models

Per seat. You pay for each agent (or each human licence sitting alongside the AI), regardless of how much it's actually used. Predictable if your usage is stable. Poor value if usage swings, since you're paying the same whether the agent handled ten conversations or ten thousand.

Per resolution. You pay only when a conversation is marked "resolved." Attractive in principle, since you're paying for outcomes rather than access. The problem is definitional. A customer might contact support two or three times about the same underlying issue, or an agent might close a conversation that looks resolved on paper but hasn't actually fixed the customer's problem. Costs become hard to predict, and disputes over what counts as a genuine resolution are common. This model has become more common through 2026, not less, with HubSpot moving its Breeze agents to per-resolution pricing layered on top of seats, and vendors including Intercom and Sierra already running significant revenue on outcome-based pricing. It's worth knowing this is the direction the market's actually moving, so a flat-fee alternative should be argued as a deliberate choice, not as "everyone else got it wrong."

Flat fee plus metered usage. A base platform fee covers the core product, with genuinely variable costs, typically voice minutes, metered on top with a bundle included and an overage alert before you're charged more. This is the model built around predictability: a business knows its floor cost every month and only sees variation on the part of the service that's genuinely variable to run.

How Copianto's pricing actually works

Copianto runs on a flat fee plus metered structure. A business chooses a monthly or annual plan (annual carries a 15% discount), and each plan includes a defined allocation of monthly credits, used for outbound activity: voice calls, WhatsApp, SMS, and email campaigns or broadcasts. Credits reset each period rather than rolling over, so the number a business sees at the start of the month is the number that matters.

If a business uses its full allocation before the period ends, it can buy additional credits as an add-on, rather than being cut off or silently charged at a different rate. Usage is visible throughout the billing period, and a low-credit alert gives a business the chance to top up before it runs short, rather than finding out after the fact.

In practice, this means the two costs that matter are the subscription plan itself and any additional credits a business chooses to buy. There's no separate per-resolution charge sitting underneath the subscription, and no ambiguity about what triggers a charge.

Questions worth asking any vendor, not just Copianto

Before comparing headline prices, it's worth getting specific answers to:

• What exactly is included in the plan, and what counts as a billable interaction?

• Are there setup or integration fees on top of the subscription?

• What happens when usage exceeds the included allowance, and how is that communicated?

• Do unused credits or included volume roll over, or reset each period?

• How are failed calls, abandoned conversations, or repeat contacts about the same issue charged?

• Are human handoffs billed separately from AI-handled interactions?

• Can you see detailed usage and billing history in real time, or only after the invoice lands?

A vendor that answers these clearly, before you ask twice, is usually a vendor whose pricing will behave the way you expect once you're actually using it.

Why predictable billing is the deliberate choice here

The market's direction, as above, is toward outcome and per-resolution pricing at the higher end. That's not being presented here as the industry making a mistake it'll eventually correct. For a large enterprise with dedicated procurement and finance resource, outcome pricing can make sense.

For a small or founder-led business, the calculation is different. The appeal of predictable billing isn't that it's always cheaper. It's that the person running the business can look at a number once a month and know, roughly, what's coming, without needing to interrogate a definition of "resolution" or reconcile a bill that moved because the product did its job well. That's a genuine trade-off, not a universal truth, and it's worth being upfront that it's a deliberate position, not a claim that outcome pricing is broken.

FAQs

What's the difference between per-resolution and flat-fee pricing?

Per-resolution charges you each time a conversation is marked resolved, which sounds fair but depends on a definition of "resolved" that varies by vendor and can be disputed. Flat-fee-plus-metered charges a base platform fee with genuinely variable costs, like voice minutes, metered separately with a bundle and overage alert.

Is per-resolution pricing becoming more or less common?

More common, not less. Several major vendors moved further toward outcome-based pricing through 2026. A predictable, flat-fee alternative is a deliberate position for buyers who value knowing their cost in advance, not a sign that outcome pricing has fallen out of favour generally.

What hidden costs should I watch for when comparing vendors?

Setup and integration fees, how repeat contacts about the same issue are charged, whether human handoffs are billed separately, and whether unused usage rolls over or resets. Ask these before signing, since they rarely appear on the headline pricing page.

Do Copianto's credits roll over if unused?

No. Credits reset each billing period rather than carrying over, which is part of what keeps the monthly cost predictable and easy to plan around.