Intercom's Fin AI Agent is sold on a simple pitch: you only pay $0.99 for every conversation it resolves. At a glance, that sounds like the cheapest way to run AI support — no per-agent seats, no flat monthly fee, you just pay for outcomes. But "pay per outcome" behaves very differently once your volume climbs from a couple hundred conversations a month into the thousands, and the sticker price rarely gets shown next to that math. Below, we work through what an Intercom Fin alternative like Callbell actually costs compared to Fin as usage scales, check the resolution-rate numbers against what support teams report in practice, and look at what a $3.6 billion acquisition by Salesforce might mean for where the product is headed next.
How per-resolution billing adds up once volume grows
Fin's pricing is built around what Intercom calls an "outcome": a resolution, a hand-off to a defined procedure, or a disqualification, each billed at $0.99, plus a $9.99 charge for a qualified sales lead. That $9.99 tier matters if you're running Fin for sales qualification rather than support — it's a different bill entirely from the $0.99 support outcomes most teams are budgeting against. Run as a standalone deployment through "Fin for platforms" — meaning outside Intercom's own helpdesk — there's a $49 monthly base fee that covers your first 50 outcomes, with every additional one billed at $0.99. Minimum monthly billing is 50 outcomes, whether you use them all or not. So if you're asking how much Intercom Fin costs in practice, the honest answer is: it depends entirely on how many conversations it touches in a given month.
Here's what that turns into once you plug in real conversation volume. The table below is a worked illustrative example built from Fin's published rate card — not an official quote for any specific business, since your actual bill depends on your mix of resolutions, hand-offs, and disqualifications, and on whether Fin is layered on top of an existing Intercom helpdesk subscription (not included here) rather than run standalone. We're comparing it to a small 3-agent team on Callbell, where the flat rate is $17 per agent per month — $51/month total — no matter how many conversations the chatbot handles.
| Monthly resolutions | Fin for platforms (illustrative) | Callbell — 3-agent team |
|---|---|---|
| 200 / month | $198/mo | $51/mo flat |
| 1,000 / month | $990/mo | $51/mo flat |
| 3,000 / month | $2,970/mo | $51/mo flat |
Even at the lowest volume in that table, the per-resolution model already costs close to four times Callbell's flat rate for a small team — and the gap only widens from there, since Fin's bill has no ceiling while Callbell's doesn't move at all as volume grows. That's the trap with usage-based AI pricing: the per-unit cost looks small in isolation, right up until it's multiplying against a number outside your control.
It isn't only visible in the math, either. Intercom Fin holds a 1.9/5 rating on Trustpilot from more than 950 reviews, with pricing transparency and slow support responses among the most repeated complaints. Callbell sits at 4.1/5 from 365 reviews, with a fair caveat of its own: reporting and analytics are fairly basic, and its automation flows are less customizable than dedicated bot-building platforms — solid for most support teams, but not the deepest toolkit in the category.
The resolution-rate gap that makes the per-outcome math worse
Marketing in this category sometimes implies autonomous resolution rates above 90%. Numbers reported for Fin in real deployments run closer to 42-50% — genuinely useful, but a meaningful gap from the headline figure (we go deeper on this pattern in our piece on what "the bot resolves it itself" really means). That gap matters more here than it would with a flat-rate tool, because of how Fin's outcomes are defined: a hand-off to a human procedure is a billable $0.99 outcome too, not only a clean resolution. So when the resolution rate sits in the low 40s, you're not saving the difference — you're paying $0.99 for the hand-off itself, and then paying an agent's time to actually finish the conversation. A lower resolution rate doesn't shrink the bill. It just changes what you're being billed for. Many vendors in this category also count a broader range of interactions as "resolved" than a support lead would recognize as fully solved — a generous definition inflates the headline number without changing what your team experiences day to day.
What the Salesforce acquisition signals
In recent corporate news, Salesforce signed a definitive agreement on June 15, 2026 to acquire Intercom — the company behind Fin — for approximately $3.6 billion, with the deal expected to close in Q4 of Salesforce's fiscal year 2027. We're not speculating here about what that means for future pricing; that's genuinely unknown at this point. What it does signal is direction: a product being folded into one of the largest enterprise CRM vendors tends to get built around large support organizations and enterprise sales motions, not around small teams looking for a predictable monthly bill.
Where each product actually lives: helpdesk suite vs. messaging inbox
Fin sits on top of Intercom's broader helpdesk suite — tickets, a help center, workflows, and now an AI layer over the whole stack, built for support organizations running a full enterprise service desk. Callbell starts from the opposite direction: one inbox for WhatsApp, Instagram, Messenger, and Telegram, with the AI chatbot answering directly inside that same inbox rather than as a feature bolted onto a ticketing system. It's built for teams whose support already happens in messaging apps, not for teams routing tickets through a help desk with chat as one channel among several. For a team fielding order updates, shipping questions, and pre-sale queries over WhatsApp, that single-inbox setup tends to mean less tool-switching and a shorter setup process than standing up Fin across a full helpdesk implementation.
So, which one should you pick?
Fin makes the most sense for large support organizations with enterprise budgets and the infrastructure to absorb variable billing — especially if conversation volume is genuinely unpredictable or low relative to headcount, and you already need Intercom's broader helpdesk feature set (soon under Salesforce's ownership). Callbell is the better fit if WhatsApp, Instagram, and Messenger are where your customers actually message you, you'd rather have the AI chatbot included in a flat per-agent price instead of metered per outcome, and you want to know your exact monthly number in a quiet month and a busy one alike. If your team is weighing an Intercom Fin alternative mainly to escape unpredictable billing, the flat-rate model is the entire point — you're trading some enterprise polish for a number you can put in a budget spreadsheet and forget about.