Intercom Fin alternatives, and what each one bills you for
Fin charges $0.99 when nobody asks a follow-up question. Eight alternatives, compared on the event that triggers a charge rather than the headline price — because that is what decides the invoice.
Every vendor in this category will quote you a price per resolution, a price per conversation, or a price per message, and all three sound equivalent. They are not. The unit is the product decision, and it is the only part of a pricing page worth reading carefully.
Intercom's Fin is the one most people compare against, so start with what its own pricing page says. A resolution is an outcome where the customer confirms Fin solved the problem or does not request more help — fin.ai/pricing — and a 24-hour period of inactivity triggers what Intercom calls an assumed resolution. That is a defensible way to run a billing system. It is also a rule under which walking away in disgust is indistinguishable from being helped.
What silence costs you
The figure below takes one month of conversations and bills it three ways, using each vendor's own published definition. The arithmetic is ours; the rules are theirs.
- Billed under this rule
- $974
- Zinx Chat, per reply
- $79
Drag the abandonment slider and watch the gap open. This is not a hypothetical failure mode: Gorgias's own platform data (2026) across more than a thousand ecommerce brands found the median wait for a human after an AI handoff is ten hours, and that 33% of AI-touched tickets never get a human reply at all. A third of the queue is people who were left waiting. Under a silence-based rule, a good share of them bill as resolutions.
A pricing model that cannot distinguish a solved problem from an abandoned one will quote you its best case and invoice you for both.
The field, in their own units
Here is everything we could verify from each vendor's published pricing, read by hand rather than scraped. Sort it by whichever column you actually care about.
| Tool | Cheapest way in | Charges by | Hands off |
|---|---|---|---|
| Zinx Chat | $29 / moStarter plan, 25 free replies first | reply | Yes |
| Tidio Lyro | $24 / moStarter, 50 Lyro conversations | conversation | Yes |
| Chatbase | $40 / moHobby, 700 message credits | message | Yes |
| Crisp | $45 / moMini, 4 seats and $5 of AI credit | conversation | Yes |
| Intercom Fin | $50 / mo50 outcomes a month, on your own helpdesk | outcome$0.99 each | Yes |
| Gorgias AI Agent | $50 / moBasic, 300 tickets a month | resolution$0.90 each | Yes |
| Zendesk AI agents | $55 / moSuite Team, one agent | resolution$1.50 each | Yes |
| SiteGPT | $59 / moStarter, 4,000 messages | message | Higher plan |
| CustomGPT.ai | $99 / moStandard, 1,000 queries | query | Higher plan |
Their own published prices, read on 2026-10-07. Entry prices are the cheapest paid way in, which for several of them is billed annually. Follow the link on each name for what it says today.
Which model suits which business
- Per resolution is best when your questions have clean endings — order status, returns, store hours. It is worst when people browse, ask three things, and leave, because each abandonment may still bill.
- Per seat plus resolutions suits a team that already lives in a helpdesk. If you are one or two people, the seat is most of the bill and the AI is a rounding error.
- Per message or per credit is the most predictable and the easiest to exhaust by accident — a stronger model can cost ten credits instead of one, and some plans simply stop answering when the cap is hit rather than charging overage.
- Per reply — what we do — charges the same whether the answer landed or not. That is honest about what we can measure, and it means a bad month costs you the same as a good one, which cuts both ways.
What to ask on the call
- What event triggers a charge, in one sentence, and where is that written down?
- If a customer abandons the chat without replying, am I billed? If they come back tomorrow and ask the same thing, is that one charge or two?
- Does a human replying in the same conversation cost extra? It should not — if the meter runs while a person types, the pricing is fighting the behaviour you want.
- What happens at the cap: overage, or silence? One is a surprise invoice and the other is a surprise outage.
- Can I see the questions it could not answer? That list is worth more than the deflection, and not every tool will give it to you.
One last thing the table cannot show. Fin's marketed average resolution rate, 76%, is higher than the 65% floor Intercom will contractually guarantee — and it describes that floor as the resolution rate of a human. Both numbers use the silence rule. When you read any vendor's resolution rate, including a number we might quote at you, the first question is not how high it is. It is who decided what counted.
If you would rather measure it yourself than take anybody's word, the measurement guide has a figure that computes three defensible deflection rates from the same inbox. They differ by more than you would like.
Common questions
- How does Intercom Fin pricing actually work?
- Fin bills $0.99 per outcome, and its pricing page defines a resolution as one where the customer confirms it or simply does not ask for more help. A 24-hour quiet period turns silence into an assumed resolution, so a customer who gives up in frustration produces the same charge as one you delighted.
- Is per-resolution pricing cheaper than per-reply pricing?
- It depends entirely on your abandonment rate. Per-resolution pricing is excellent when most conversations genuinely end well, and punishing when they do not, because the billing rule cannot tell the two apart. Per-reply pricing charges the same either way, which is worse at low volume and more predictable at high volume.
- Do I need an Intercom subscription to use Fin?
- Fin runs on your existing helpdesk, including Zendesk, so you are not forced onto Intercom's own inbox. But running it inside Intercom adds a seat from $29 a month, and the cheapest realistic way in is roughly $49.50 for 50 outcomes.
- Is Fin's advertised resolution rate reliable?
- Yes, in both directions. Fin AI's own site claims an average of 76% across 12,000-plus customers, while Intercom's commercial guarantee sets a floor of 65% and calls that the human resolution rate. Neither figure comes with a published methodology, and both use the silence-based definition of a resolution.