Choose Intercom Fin when
Teams already on Intercom, or wanting one agent that spans voice, chat, email and social with the vendor managing the models.
RAUM AI vs Intercom Fin
Intercom Fin or a chatbot billed per conversation on your own AI account?
Fin charges when it produces an outcome and runs on Intercom's own models. RAUM charges per conversation and runs on the provider account you already pay. The right answer depends on your resolution rate and your appetite for owning the model relationship.
Last reviewed: September 2, 2026
14-day free trial. No credit card required.
Decision guide
Which approach fits best?
Choose RAUM AI when
Teams answering website questions from documented procedures who want predictable per-conversation cost and their own provider account.
Neutral considerations
What to validate before choosing
Scope
Fin is positioned across many channels and helpdesks. RAUM Chat is a website channel on the same engine as the Zendesk product, so compare like for like before comparing price.
Two different meters
Per outcome and per conversation are not convertible without knowing your resolution rate. Model both against your own numbers rather than comparing headline rates.
Model ownership
Intercom states Fin runs on its own Apex models. RAUM has no models of its own and answers with credentials you supply.
Side by side
RAUM AI vs Intercom Fin
Competitor cells use the official sources listed below. “Not publicly documented” means we did not find a current official statement; it is not a claim that the capability is absent.
| Decision area | Intercom Fin | RAUM AI |
|---|---|---|
| Metering unit | Per outcome, charged once per conversation[1] | Per conversation, charged once per conversation id |
| Published rate | $0.99 per outcome, plus seats at $29, $85 and $132 per seat/mo[1] | $49, $149 or $399 per month including 500, 2,000 or 6,000 conversations; no seats |
| AI-provider credentials | Vendor-operated: Intercom states Fin runs on its Apex models[2] | You connect your own OpenAI, Anthropic or Google account |
| Channel coverage | Positioned across voice, chat, email, Slack and social[2] | Website chat, plus the Zendesk ticket channel as a separate product |
| Helpdesk independence | Native to Intercom and documented as working with other helpdesks[2] | Independent of the helpdesk; hands off by email, webhook or Zendesk ticket |
| Model spend | Included in the per-outcome price | Billed by your provider at your rate, with no markup from us |
| Live web knowledge | Confirm crawling, exact live-page access, and freshness behavior in the cited product documentation | Website crawling, exact live-page lookups, dynamic URL variants, and durable last-good snapshots |
| Workflow review controls | Confirm current builder, approval, review, and simulation controls in the cited documentation | Guided connected workflows, per-finding AI Review, ticket directives, and Simulation Mode |
| Zendesk onboarding | Product-specific installation and setup | Guided Marketplace installation with automatic webhook and trigger setup |
In depth
What the choice actually turns on
Per outcome and per conversation are the same bill at one specific resolution rate
Charging only for outcomes sounds strictly better, and for a low-deflection queue it is: you pay nothing for the conversations the bot could not help with. The crossover moves as your content improves, though, because a per-outcome meter charges you more precisely when the bot gets better, while a per-conversation meter does not. Neither model is dishonest; they simply reward different things. Take your real conversation volume and your real resolution rate, and compute both — the arithmetic answers this far better than either vendor's headline number.
Owning the provider relationship is the part you cannot add later
Everything else in a chatbot evaluation can be changed after launch — content, tone, escalation rules, which pages it runs on. Where the inference happens cannot, because it determines who holds the contract, who sees the data, and whose rate card your unit cost tracks. RAUM answers with your own provider credentials, so your customer data flows from your site to a provider you have your own agreement with, and the model bill is one you can renegotiate or move. If that does not matter to your team, a vendor-managed model is genuinely less work.
Setup and migration
A practical evaluation sequence
- 1
Pull the last quarter's Fin outcome count and total conversation count. The ratio between them is what makes the two pricing models comparable at all.
- 2
Separate what Fin does on your website from what it does on other channels, and scope the comparison to the website only.
- 3
Connect a provider account and run RAUM Chat on a subset of pages behind an origin allowlist before moving any traffic.
FAQ
Questions teams ask
Is RAUM cheaper than Fin?
It depends entirely on your resolution rate, and anyone who answers this without asking for that number is guessing. Fin bills per outcome; RAUM bills per conversation whether or not it resolved. Take last quarter's conversations and outcomes and compute both — and remember to add your own provider's model spend on the RAUM side, which is real even though we do not mark it up.
Does RAUM Chat work on voice or social?
No. It is a website channel. The Zendesk product covers email and the ticket queue as a separate product on the same engine. If you need one agent across voice and social, that is a genuine reason to prefer a broader platform.
Do I need Intercom or any particular helpdesk to use RAUM Chat?
No. It runs on your own site and hands conversations off by email, a signed webhook, or a Zendesk ticket. Zendesk is one destination, not a requirement.
Which models does RAUM use?
Whichever you connect. RAUM has no models of its own — you supply an OpenAI, Anthropic or Google account and choose the model per purpose, so you can move to a cheaper or stronger model without asking us.
Evidence
Official sources
- [1] Intercom pricing
Seat prices and the $0.99-per-outcome Fin rate; verify at purchase.
- [2] Fin — the highest performing Customer Agent
Channel coverage, helpdesk compatibility, and the statement that Fin runs on Apex models.
Product packaging and availability can change. Recheck these official pages and obtain a vendor quote before making a purchase decision.
Keep exploring
Related resources
Evaluate RAUM AI with your real Zendesk workflows
Install from the Marketplace, connect a supported provider, and verify representative tickets without sending customer replies.
14-day free trial. No credit card required.