Fin is Intercom's AI customer-service agent that resolves support conversations on its own — answering from your help center and past tickets, taking actions, and handing off to a human only when needed — built for support and CX teams at companies of any size that want to cut ticket volume without losing quality.
Who it's for
Fin fits customer-support organizations that already run (or are willing to run) a modern helpdesk and want an agent that resolves real tickets, not just a FAQ widget. It's used by everything from scaling SaaS startups to large support teams handling thousands of monthly conversations. Because Fin bills per resolved conversation rather than per seat, it particularly suits teams that want costs to track outcomes: you pay for tickets it actually closes, not for the agent sitting idle. Teams wanting a fully free, unlimited chatbot should look elsewhere, since Fin has no free tier.
How it works
Fin reads your knowledge base, macros and historical tickets, then answers customer questions in natural language and can take actions — looking up an order, updating a record, triggering a workflow — through connected apps and APIs. It runs semi-autonomously: it resolves what it's confident about and escalates ambiguous or sensitive cases to a human agent, keeping a person in the loop for the conversations that need judgment. Fin is native to Intercom's helpdesk but is also available "over API," so teams running Zendesk, Salesforce or a custom support stack can plug Fin's resolution engine into their existing tools rather than migrating their whole helpdesk.
Pricing
Fin uses outcome-based pricing: you pay roughly $0.99 per resolution, meaning you're billed only for conversations Fin actually resolves rather than for every message or seat. There is no free tier. This model rewards a well-maintained knowledge base — cleaner content means more resolutions and a more predictable bill — but always check Intercom's current pricing page, since resolution pricing and any bundled plans can change.
Strengths and trade-offs
Fin's biggest strength is that its pricing is tied directly to value delivered, which is unusually transparent for AI support tools, and it is SOC 2 audited with an API for teams that don't want to fully adopt Intercom. The semi-autonomous escalation model helps avoid the classic AI-support failure mode of confidently answering wrong. The trade-offs: there's no free tier to test cheaply, per-resolution costs can add up at very high volume, and Intercom doesn't publish a dedicated EU-hosting region, so organizations with strict data-residency requirements should verify hosting details before committing. For support teams focused on measurable deflection with a human safety net, Fin is one of the most established options on the market.