Intercom Pricing: The Ultimate Guide [2026]

If you're shopping for customer support software, Intercom is one of three or four names you'll see on every comparison page. It's also one of the few platforms that charges in two completely different ways at the same time: per seat for your humans, per resolution for its AI.

That dual-pricing model makes "how much does Intercom cost?" a harder question than it sounds. In this guide, we'll break down every Intercom plan, what the $0.99-per-resolution Fin AI Agent actually adds to your bill, the add-ons most articles skip, and how Intercom compares to the alternatives serious support teams shortlist next to it.

TL;DR: Intercom's 2026 Pricing at a Glance

Three things to notice before reading the rest of this guide. Seat pricing is per agent per month, billed annually for the best rate. Fin AI Agent is billed separately on top of any seat plan, or as a standalone product if you only want AI. And the add-on stack (Copilot, Pro, Proactive Support Plus, Phone) routinely doubles the headline cost.

What Intercom Actually Costs

Intercom retired its always-free Starter plan around late 2023. As of 2026, every paying customer lands in one of three seat-based tiers. A 14-day free trial is available without a credit card, but there's no permanent free tier.

Essential Plan: $29/Seat/Month (Annual)

The cheapest path into Intercom. At $29 per seat per year (or $39 monthly), Essential includes the basics most teams associate with Intercom: the Messenger widget, shared inbox, ticketing, public help center, simple automations, pre-built reports, and Slack integration. Fin AI Agent is available on top at $0.99 per resolution.

Best for: solo founders, startup support teams of three to ten agents, or any team that wants to get on Intercom without committing to workflows or multi-team setups.

Advanced Plan: $85/Seat/Month (Annual)

The middle tier and the most common choice for mid-market support teams. Advanced costs $85 per seat per year ($99 monthly) and adds custom workflows, multiple team inboxes, round-robin assignment, multilingual help center, private help center, ticket portals, custom reports, and Salesforce + Marketo integrations.

It also bundles 20 free Lite Seats. Lite Seats can read and write internal notes, but can't respond to customers directly. If you have engineers or product managers who occasionally jump into tickets, that 20-seat allowance can shave real money off the bill.

Expert Plan: $132/Seat/Month (Annual)

The enterprise tier. At $132 per seat per year ($139 monthly), Expert adds the controls regulated industries actually need: SSO and identity management, HIPAA support, SLAs, multibrand messenger and help center, workload management, team office hours, team reply-time settings, extended API limits, and a real-time dashboard. It also includes 50 free Lite Seats.

If you're in healthcare, financial services, or any regulated industry, Expert is the only tier that gives you HIPAA support and SSO. Below it, you're either paying for those features piecemeal or you're not getting them at all.

The $0.99-Per-Resolution Fin AI Agent (What Outcome-Based Pricing Actually Means)

Fin AI Agent is Intercom's most-discussed product in 2026 and the reason "intercom pricing" still surfaces fresh threads on Reddit and LinkedIn. The model is simple: when Fin successfully resolves a customer's conversation without a human agent, Intercom charges $0.99. When it can't, you pay nothing.

That sounds clean, but a few details matter when you do the math.

What counts as a resolution. Intercom defines a resolution as Fin closing a conversation in a way that either (a) the customer confirms their issue is solved, (b) the customer is inactive after Fin's response and Fin's reply is rated as helpful, or (c) Fin successfully routes the conversation to the right outcome (knowledge article, form submission, ticket creation). Conversations Fin escalates to a human don't count.

Per-month minimums apply. Standalone Fin AI Agent (without the seat plans) has a minimum monthly commitment, even if your actual resolution volume is below it. For seat-plan customers, Fin's resolution charges are added on top of the seat fees.

Standalone or bundled. You can buy Fin AI Agent without any Intercom seats and plug it into your existing helpdesk (including Salesforce). That's a deliberate move: Intercom wants Fin to compete on its own merits against pay-per-resolution players like Decagon or Sierra without forcing you to swap your help desk.

For teams currently paying $5–$15 per ticket for human-handled support, $0.99 per resolution is a meaningful drop. For teams where most "resolutions" are FAQ-tier deflection that a knowledge base could handle, the math gets uglier fast.

Other Add-Ons That Will Affect Your Bill

Most pricing articles stop after Essential / Advanced / Expert. The add-ons are where Intercom bills actually balloon.

Fin AI Copilot at $29 per agent per month. A different product from Fin AI Agent. Copilot rides shotgun with your human agents, drafting replies, summarizing conversations, and surfacing help articles. Priced per seat (not per resolution) for unlimited use.

Pro add-on at $99 per month. Includes analysis of 1,000 conversations per month. Adds AI-powered conversation insights, custom AI features, and deeper analytics. Beyond 1,000 conversations, you're paying overages.

Proactive Support Plus at $99 per month. Includes 500 outbound messages per month. Adds product tours, banners, multi-step messages, and outbound campaigns. Like Pro, anything beyond the cap costs extra.

Intercom Phone (usage-based). No per-seat fee, but you pay for phone numbers (free in some countries, up to $275 for Chile toll-free) plus inbound minutes ($0.012–$0.596 depending on country and line type) and outbound minutes (varies by destination). Messenger Calls cost $0.03 per minute plus $0.03 per recorded minute.

SMS, WhatsApp, email campaigns. All pay-as-you-go, billed per message.

Stacking the math for a typical mid-market team: ten Advanced seats ($850/mo) + Copilot for all ten ($290/mo) + Pro ($99/mo) + Proactive Support Plus ($99/mo) + Fin AI Agent at 500 resolutions/mo ($495/mo) = $1,833/month, or roughly $22,000/year before any phone usage. The headline "$85 per seat" number describes about 46% of what you'll actually pay.

What Intercom Doesn't Tell You About Annual Vs Monthly

The price gap between annual and monthly billing is steeper than the pricing page makes obvious. Essential jumps from $29 (annual) to $39 (monthly), a 34% markup. Advanced moves from $85 to $99 (+16%). Expert from $132 to $139 (+5%).

If you're confident you'll stick with Intercom for a year, annual billing is the right call. If you're testing the platform during a 14-day trial and want to keep flexibility, monthly billing buys you that option at a real premium. Most teams sign annual on second renewal once they've validated the platform.

There's also the Lite Seat trap. Lite Seats sound free, but they're capped per tier (20 on Advanced, 50 on Expert), and they can't actually respond to customers. If you have 30 engineers who occasionally answer support questions and you're on Advanced, you'll need to upgrade or pay for full seats once you cross 20.

Intercom Vs. Alternatives at a Glance

Most teams comparing Intercom also evaluate two or three of these alternatives. Here's how the pricing models stack up:

Two patterns to notice. The legacy CX platforms (Zendesk, LivePerson, Cognigy) are layering outcome-based AI pricing on top of seat-based human pricing. The AI-native players (Sierra, Decagon, Fin standalone) are charging per resolution without a seat-based human layer. Voiceflow sits in a third category: a build-and-run platform where the AI logic is yours, and the marginal cost of a resolution is whatever the underlying LLM provider charges, not a vendor markup.

Is Intercom Worth It?

For mid-market support teams whose conversations are mostly chat, mostly English, and mostly tier-1 deflectable, Intercom in 2026 is a reasonable default. The Messenger widget is still best-in-class, Fin works well on common ticket types, and the new outcome-based pricing aligns your AI spend with actual value created.

For teams that hit any of the following, the answer gets harder:

• You need real voice or phone coverage. Intercom Phone is a thin layer; voice still feels like a chat company's add-on. Native voice platforms like Voiceflow, Sierra, or LivePerson handle voice as a first-class channel.
• You want to choose the underlying LLM. Fin runs on OpenAI plus a thin Anthropic layer. If you have an existing Bedrock commitment, a preferred GLM provider, or want to swap models per use case, Intercom doesn't expose that.
• You're at enterprise scale with regulatory needs beyond HIPAA. Expert covers HIPAA + SSO, but anything more bespoke (SOC 2 Type 2 segregation, EU residency, PII tokenization on specific fields) tends to push teams toward custom-built or enterprise platforms with explicit security controls.
• Your "resolution volume" is mostly FAQ-tier. If most of what Fin would resolve is a customer asking the same five questions, building those flows yourself on a platform you own is often 5–10× cheaper at scale.

The historical complaints about Intercom (laggy search, confusing navigation, slow support response from Intercom's own team, expensive Product Tours add-on, limited mobile compatibility for some add-ons) have mostly held through the 2026 redesign. They're not dealbreakers, but they're not gone either.

Building AI Agents That Plug Into Intercom (Without Paying Per Resolution)

If you like Intercom's Messenger and ticketing but balk at $0.99 per AI resolution, you can run your own AI agent on top of Intercom's infrastructure. That's what most Voiceflow customers do.

Voiceflow is a platform for building, launching, and scaling AI agents across chat and voice channels. Teams like Turo, StubHub International, Sanlam Studios, and Trilogy use it for production agents that handle support, sales, and operations.

Five things Voiceflow does differently from Fin:

1. Pay once per build, not per resolution. You design your agent's logic once and run it on Voiceflow's platform. The marginal cost of an additional resolution is the underlying LLM call (often pennies), not a $0.99 vendor markup.
2. Native voice and phone. Voice + phone are first-class channels with deterministic Workflows + LLM Playbooks + Tools (Function/API/MCP). Includes call_forward for human handoff, dtmf for keypad input, and multi-provider STT/TTS. See AI call center agent and voicebot for examples.
3. Model-agnostic by design. Run agents on OpenAI, Anthropic, Google, Bedrock, GLM, or Groq. Swap models per workflow without re-architecting. Fin doesn't expose model choice.
4. Knowledge Base built in. Chunked semantic search, OpenAI embeddings, REST filter operators, optional LLM synthesis. Deflects tier-1 tickets without paying per resolution.
5. Production primitives included. Evaluations (pre-launch agent tests), Observability (conversation-level visibility, debugging), Environments (dev / staging / production). SOC 2 Type 2 + PII masking baked in.

Most Voiceflow + Intercom teams keep Intercom for the inbox + Messenger and route AI conversations to a Voiceflow agent via Intercom's API. You get the Intercom UX your agents are used to, and you don't pay per resolution to a vendor for AI work you can run cheaper yourself. See AI customer service agent and customer service chatbot for the broader category, plus customer service automation for category framing.

If you're a CX leader thinking about ROI math on this swap, our AI customer service ROI guide and the agentic AI in the contact center 2026 landscape explore the numbers in detail. Curious about the broader AI chatbot landscape Intercom now competes in? See our best AI chatbot roundup.