Intercom Fin Pricing: The Per-Resolution Cost Breakdown (2026)

If you want the full seat-plan rundown (Essential/Advanced/Expert and what each tier unlocks), that lives in Intercom pricing explained; if you want what Fin actually does under the hood, see the Fin AI explainer. This post stays narrowly on the cost model. Every figure is verified against fin.ai/pricing and the Fin Help Center as of June 2026 — Fin revises pricing periodically, so confirm the live numbers in your own account before they anchor a budget.

Update — 2026: On June 15, 2026, Salesforce signed a definitive agreement to acquire Fin (formerly Intercom) for roughly $3.6 billion, with the Fin AI agent folding into Salesforce's Agentforce platform; close is expected around Q4 of Salesforce's fiscal 2027. The $0.99 model below is unchanged today, but pricing and packaging could shift under Agentforce — more at the end.

The model in one line

Fin doesn't charge per seat, per message, or per conversation. It charges **per billable outcome — $0.99 each — and at most once per conversation**, on top of whatever you already pay for your help desk seats (fin.ai/pricing). So the formula is:

Your Fin bill ≈ (billable outcomes × $0.99) + your seat plan + any add-ons — with a hard floor of a 50-outcome/month minimum (~$49).

The thing that makes this hard to budget is the first term. "Billable outcomes" is not the same as "tickets," "conversations," or even "problems solved." Getting that count right is the whole game.

What counts as a billable outcome

Fin recognizes four outcome types. Three cost $0.99; one costs ten times that (Fin pricing: Outcomes):

For a typical support team, resolutions are ~all of the bill — handoffs only count when you configure a Procedure to end in one, disqualifications and qualifications only apply if you run Fin on sales/lead flows, and the $9.99 qualification is the one line item that can quietly distort a sales-heavy bill. The free row matters just as much: if Fin can't solve it and simply punts to a human, you pay nothing. You broadly pay when Fin does the work, not when it gives up.

Confirmed vs. assumed resolution — and the 24-hour window

This is the single most important mechanic in Fin pricing, and the one teams under-model. A "resolution" is counted two ways (fin.ai help — outcomes):

• Confirmed resolution — the customer explicitly signals it helped ("Ok, thanks", "that worked").
• Assumed resolution — the customer disengages for 24 hours after Fin's last answer without asking for more help. After that window with no reply, Fin marks it resolved and bills $0.99.

The assumed case is the wrinkle. A customer who reads Fin's reply and wanders off — whether they were genuinely helped or quietly gave up — becomes a billable resolution once 24 hours pass. It's a defensible default (most people who go quiet were probably helped), but it means your billable count can run higher than the number of problems actually solved. The practical consequence: your Fin invoice is not a clean proxy for CSAT. Audit assumed-resolution rate against satisfaction in month one before you trust the deflection math.

The fairness mechanics (these are genuinely buyer-friendly)

To Fin's credit, the meter has guardrails that beat most usage pricing:

• One outcome per conversation, no matter how many questions Fin answers or actions it takes.
• No charge if the customer asks for a human, or if a configured Fin Procedure fails to complete.
• Refund on reopen — if a conversation marked resolved is reopened later by the customer seeking more help, that resolution is deducted (even across billing periods).

The 50-outcome minimum, no rollover, no volume break

Three structural details shape the floor and the ceiling of your bill (fin.ai/help — pricing: outcomes):

• Minimum commitment of 50 outcomes/month — the base plan is ~$49/month and includes 50 resolutions. Below 50, you still pay for 50. This is the practical floor.
• No rollover. Unused monthly resolutions are lost — they don't bank toward a busy month. If you buy/commit ahead, you're paying for a monthly allotment, not a pool.
• No advertised volume discount. The public rate is a flat $0.99 whether you do 60 outcomes or 60,000. (Enterprise commitments may negotiate — but nothing on the public page promises a tier break, so don't assume one.)
• There's a 14-day free trial with unlimited outcomes and no card — the right way to get a real resolution count before committing.

The flat rate is the headline cost characteristic: Fin's pricing scales linearly with volume forever. That's cheap when volume is low and deflection is modest, and it's the line item that balloons fastest as you grow — the opposite shape of seat pricing, which flattens out.

How Fin stacks on top of Intercom seats

Fin is almost never your only Intercom cost. If you run it on Intercom, you also pay per-seat for the platform underneath (intercom.com/pricing) — billed annually, per seat/month:

• Essential — $29/seat/mo
• Advanced — $85/seat/mo (most growing teams land here)
• Expert — $132/seat/mo

Plus optional add-ons that also sit on top: Copilot (agent-side AI, ~$29/agent/mo annual), Proactive Support Plus ($99/mo incl. 500 messages), and pay-as-you-go SMS/WhatsApp/phone. None of those are included in the $0.99. So the real all-in is seats + Fin outcomes + add-ons — and as the scenarios below show, for a busy team Fin alone routinely outweighs the seats.

Worked cost scenarios

Here's the math that actually matters. All three use annual seat rates, count only $0.99 resolutions (no sales qualifications), and ignore Copilot/Proactive so the Fin-vs-seats ratio is clean. Resolution volumes are illustrative — plug in your own.

Scenario A — Small team: 3 seats, 500 resolutions/mo

Even at small scale, Fin is ~66% of the bill and the "all-in cost per resolution" is ~$1.50 once you fold in the seats those resolutions still require.

Scenario B — Mid-market: 10 seats, 2,000 resolutions/mo

Now Fin is ~70% of the bill. The "$850/month, 10 seats" sticker becomes a ~$34k/year run-rate — and that's before Copilot (add ~$290/mo for 10 agents). All-in cost per resolution: ~$1.42.

Scenario C — High volume: 25 seats, 10,000 resolutions/mo

At volume, Fin is ~75% of the bill — roughly 3× the seat cost — and because there's no volume discount, that ratio only grows. All-in cost per resolution: ~$1.32. This is where the flat per-outcome rate stops feeling cheap: every additional 1,000 resolutions adds a flat $990/month, indefinitely.

The pattern across all three: your headcount sets the floor; your resolution volume sets the bill. Two 10-seat teams can differ by $20k+/year purely on Fin volume.

How to estimate your own Fin bill

A quick back-of-envelope that gets you within striking distance:

1. Start with monthly AI-eligible conversations — chats/emails Fin will actually attempt (not your total ticket count).
2. Apply an honest resolution rate. Fin/Salesforce cite ~76% autonomous handling and teams reaching the high 80s–90s, but real-world deflection commonly lands 50–80% and tracks your knowledge-base quality. Use a conservative number first.
3. Add the assumed-resolution effect. Expect billable resolutions to run somewhat above "truly solved," since 24-hour disengagements count. Don't shave this off.
4. Compute outcomes × $0.99, then enforce the 50-outcome floor.
5. Add seats (seats × plan rate) and any add-ons (Copilot ~$29/seat, Proactive $99, channels).
6. Add sales outcomes if relevant — qualifications at $9.99 each can move the number fast.

Roughly: All-in = max(50, conversations × resolution rate) × $0.99 + (seats × seat price) + add-ons. Run the 14-day trial to replace the resolution-rate guess with a real figure before you commit.

Where the cost surprises teams

The line items that show up on the invoice but not in the mental model:

• Assumed resolutions inflate the count. The 24-hour-silence rule bills conversations the customer may not have been happy with. This is the #1 gap between estimate and invoice.
• **Fin usually isn't the whole bill — it's the bigger part.** Seats are the floor; Fin is the variable that overtakes them. Budgeting from the seat sticker undershoots badly.
• No rollover. A slow month doesn't credit a busy one; you can't smooth spend across the year.
• No volume discount on the public rate. Costs scale linearly forever — there's no economy-of-scale relief baked into the list price.
• The $9.99 qualification. Harmless for pure support, meaningful if you point Fin at lead-gen — it's 10× a normal outcome.
• Copilot is a separate AI line. Wanting agent-side and customer-side AI means two AI bills, not one.
• Channels are metered. SMS/WhatsApp/phone are pay-as-you-go on top.

A note on billing philosophy (and where Macha differs)

A quick, honest aside, since this is a Macha guide. **Macha is an AI agent layer that runs on top of Zendesk and Freshdesk only — it does not integrate with Intercom**, so it's not a drop-in swap for Fin. The reason it's worth a mention here is purely the unit of pricing, which is a real fork in the road.

Fin bills per resolution — one tidy outcome per conversation, including assumed ones. Macha bills per AI action — any automated step it takes (triaging, tagging, routing, drafting, resolving) — because most useful automation isn't a single clean "resolution," it's work done across a ticket's life. Neither shape is universally cheaper: per-resolution is simple and predictable when your work really is discrete solved chats; per-action maps better when an agent does several small things per ticket and you'd rather pay for the work than guess at an "outcome." If your help desk is Zendesk or Freshdesk and the per-resolution math is what worries you, it's worth modeling both — see Macha on Zendesk, or 7-day free trial, no credit card required to run it against your own tickets. We go deeper on the trade-off in Fin vs. an AI agent layer. If you're committed to Intercom, Fin is the native path and this doesn't apply — but the contrast is useful either way.

Update — 2026: the Salesforce acquisition

On June 15, 2026, Salesforce signed a definitive agreement to acquire Fin (formerly Intercom) for ~$3.6 billion, folding Fin's AI agent into Agentforce (Salesforce newsroom, TechCrunch). Salesforce positions Fin as the packaged, fast-to-deploy option alongside Agentforce's deeper enterprise build, and cites Fin handling ~76% of queries autonomously. The deal is expected to close around Q4 of Salesforce's fiscal 2027.

For pricing, practically: nothing changes today — the $0.99 model stands. But packaging and pricing frequently shift after an acquisition closes, and Fin's per-outcome model could be repackaged under Agentforce's licensing. If you're signing multi-year, build in a re-evaluation clause.

Frequently asked questions

How much does Intercom Fin cost? Fin is billed $0.99 per outcome, charged at most once per conversation, with a 50-outcome/month minimum (~$49 base). On Intercom, that stacks on top of per-seat plans ($29/$85/$132 per seat/month, billed annually). For a 10-seat team handling ~2,000 resolutions/month, the all-in run-rate is roughly $2,800/month (~$34k/year) — and Fin is the majority of it.

What counts as a Fin resolution? Either a confirmed resolution (the customer says it helped) or an assumed resolution (the customer disengages for 24 hours after Fin's last answer without asking for more help). Both bill $0.99. You're not charged if the customer asks for a human or a configured Procedure fails, and a resolution is refunded if the customer reopens the conversation later.

Is the $0.99 the same for every outcome? Resolutions, procedure handoffs, and disqualifications are all $0.99. A sales qualification is $9.99. Escalations to a human with no resolution or configured handoff are free.

Is there a Fin cost calculator or volume discount? The public rate is a flat $0.99 per outcome with no advertised volume discount and no rollover of unused monthly outcomes. The most reliable "calculator" is: max(50, conversations × resolution rate) × $0.99 + seats + add-ons. Run the 14-day free trial to get a real resolution count first.

Does Fin pricing include Intercom seats? No. Fin's per-outcome fee is separate from — and on top of — Intercom's per-seat plans and any add-ons like Copilot (~$29/agent/mo) or Proactive Support Plus ($99/mo). For the full seat breakdown, see Intercom pricing explained.

Will Fin pricing change after the Salesforce acquisition? Not immediately. Salesforce announced the ~$3.6B acquisition on June 15, 2026, with Fin folding into Agentforce and close expected around Q4 of Salesforce's fiscal 2027. The $0.99 model is unchanged today, but expect possible repackaging under Agentforce over time.

The bottom line

Intercom Fin pricing is simple to state and easy to under-budget. The $0.99-per-outcome rate is real and the fairness mechanics (one charge per conversation, free human escalation, refund on reopen) are genuinely good. But the bill is governed by your resolution count, not your headcount — and that count includes 24-hour "assumed" resolutions, carries a 50-outcome floor, doesn't roll over, and gets no volume discount, all stacked on per-seat plans underneath. In our worked scenarios, Fin runs 66–75% of the all-in cost, turning an "$850/month" seat plan into a ~$34k/year reality at mid-market volume. The honest advice: model your conversation volume and an honest resolution rate first, enforce the floor, add seats and add-ons, and run the 14-day trial to replace the guesswork. Then you'll land close to your actual invoice — and you can weigh whether per-resolution is the billing shape you want, or whether a per-action model fits your work better. For the wider context, see Intercom pricing explained, the Fin AI explainer, and Fin vs. an AI agent layer.

Fin pricing verified against fin.ai/pricing and the Fin Help Center (articles 13975800 and 8205718), June 2026. Fin revises pricing and packaging periodically — and is mid-acquisition by Salesforce — so confirm exact numbers in your own account before relying on them. Worked-scenario volumes are illustrative.