Chatbot pricing pages are designed to be compared badly. One vendor charges $0.99 per "resolution" — sounds like pocket change until you're doing 500 of them a month. The other charges a flat $29 plus a $39 AI add-on — sounds predictable until usage tiers kick in. We modeled both with real numbers so you don't learn the difference from your invoice.
The two pricing religions: flat-rate vs. per-resolution
Flat-rate pricing (Tidio's model) charges you for the platform and a quota of AI conversations: predictable, but the quota is the trap — blow past it and you're buying add-on packs or upgrading tiers mid-month. Per-resolution pricing (Intercom Fin's model) charges only for conversations the AI actually resolves: elegant in theory, but your bill scales linearly with success, which means the better it works, the more you pay. Neither is a scam; they're just optimized for different volumes.
Tidio/Lyro: what $29 + $39/month actually buys
Tidio's base plans start around $29/month for the live-chat platform (team inbox, widgets, basic automation). The AI agent, Lyro, is an add-on from around $39/month for 100 AI conversations. So the real entry price for an AI chatbot is ~$68/month, not $29 — the pricing page's headline number buys you the chat window, not the brain.
What $68 gets you: an AI that answers from your website content and FAQs, hands off to humans with full transcripts, and works fine for a business fielding 100–300 chats a month. Where quotas bite: exceed 100 AI conversations and you're buying more at roughly $0.40–$0.50 per extra conversation, or stepping up a tier. A store doing 400 AI chats a month is realistically paying $100–$140, not $68. Model your volume first — check last quarter's chat logs before you pick the tier.
Honest limitation: Lyro is only as good as the knowledge base you feed it. Garbage content in, confident wrong answers out. Budget an afternoon writing real answers to your top 10 questions, and set the two-strike handoff rule: if the bot can't resolve it in two tries, a human takes over.
Intercom Fin: $0.99/resolution modeled at 100, 500, and 2,000 resolutions
Fin charges $0.99 per resolution — a conversation the AI handles end-to-end without human help. No platform fee games, no quotas: the math is brutally simple.
| Resolutions/month | Monthly cost | What that business looks like |
|---|---|---|
| 100 | $99 | Small service business, FAQ-heavy chats |
| 500 | $495 | Growing e-commerce store, order-status volume |
| 2,000 | $1,980 | High-volume support operation |
At 100 resolutions, Fin is merely pricey. At 500, it's a car payment. At 2,000, it's a part-time employee — which is exactly the comparison to make, because that's what it's replacing. Note the definition matters: a "resolution" means the AI finished the job. Chats that escalate to your team don't count, which is fair — but it also means your bill rewards the AI for keeping humans out of the loop, for better or worse.
Honest limitation: Fin is the best conversational AI of the bunch and the most expensive at scale. For a 10-person business, you're paying enterprise-grade AI prices. The quality gap over Lyro is real but narrow for simple FAQ workloads — you're paying for sophistication you may not use.
The crossover chart: at what volume each option wins
- Under ~150 AI conversations/month: Tidio wins on price ($68 vs. ~$150 for Fin). The quality gap doesn't matter at FAQ volume.
- ~150–350/month: the gray zone. Tidio with add-on packs runs $100–$180; Fin runs $150–$350. Pick on features, not price — Fin's handoffs and analytics are better; Tidio's flat-ish cost is easier to budget.
- 350+/month: Fin pulls ahead on capability but the bill stings. This is where you should be asking whether a chatbot is even the right tool (see below) — or negotiating an annual plan, which both vendors discount 15–20%.
One more variable: resolution rate. If your chats are complex and only 40% resolve without a human, Fin's effective cost per chat is $0.40 — suddenly cheap. If 90% resolve, you're paying $0.89 per chat. Know your complexity before you model.
The third option: when a simple FAQ bot (or better docs) beats both
Here's the heretical take: if 80% of your chats are the same five questions, you might not need AI at all. A well-built FAQ page, an order-tracking page that actually works, and a rules-based bot with buttons ("Track my order" / "Book a service" / "Talk to a human") can deflect the same volume for $0–$20/month. AI chatbots earn their keep on the long tail — the weird, phrased-ten-different-ways questions. If your long tail is short, save the money and fix your help docs. That's a $0 solution with a one-afternoon setup cost.
Verdict by chat volume tier
- Under 150 chats/month: Tidio + Lyro. ~$68/month, plenty capable, easy budget line.
- 150–500 chats/month: Tidio if your questions are simple and repetitive; Fin if they're varied and you value handoff quality. Run a 30-day trial of both with real transcripts — this tier is close enough that your actual conversations decide.
- 500+ chats/month: Fin, but negotiate annual pricing and audit monthly. At $495+/month, a quarterly "is this still worth it" review is mandatory — compare against the cost of the part-time support hire it's replacing (~$800–$1,200/month loaded). If the bot isn't clearly cheaper, it isn't working.
And re-check pricing quarterly. Chatbot pricing moves faster than any other category we cover — stale numbers kill trust in this post faster than anywhere else, so we refresh these tables every three months.
FAQ
What exactly counts as a "resolution" on Intercom Fin?
A conversation the AI handles start to finish without a human teammate stepping in. If the customer asks to talk to a person, or the AI escalates, it's not billed as a resolution. Simple questions resolved cleanly are the cheapest possible outcome — which aligns the pricing with what you want.
Can I start on Tidio's free tier and upgrade later?
Yes, and you should. The free tier covers live chat without the AI agent — use it to learn your real chat volume for a month, then buy the Lyro add-on sized to actual numbers instead of guesses.
Will an AI chatbot hurt customer satisfaction?
Only if it's a dead end. Bots that resolve routine questions and hand off cleanly to humans improve satisfaction scores; bots that trap people in loops destroy them. Configure the escalation path before you configure anything clever.
How do these compare to just hiring a VA for chat?
A part-time VA at $8–$12/hour covering 20 hours a week costs ~$700–$1,000/month and handles nuance better. The chatbot wins under ~300 chats/month on cost; above that, compare carefully — and consider the hybrid: bot for nights and weekends, human for business hours.