Tidio Lyro pricing vs Fin and Gorgias: the real cost per solved question

Tidio bills Lyro by the conversation the AI picks up. Intercom's Fin, Help Scout and Re:amaze bill by the problem the AI actually solves. Those two rates are not comparable, and the gap between them is your resolution rate. At Lyro's own advertised 67 percent, Tidio's entry rung of roughly 0.65 per conversation works out near 0.97 per solved question, which lands almost exactly on Fin's $0.99 per resolution. The sticker looks 34 percent cheaper. The arithmetic says it is not.

Last updated August 2026. Every rate below was read first-hand from the vendor's own pricing page on 20 August 2026.

What each vendor charges, and what the unit actually is

The single most useful thing you can do before signing an AI support contract is write down the unit. Not the number, the unit. Six vendors publish six different ones, and a per-conversation rate next to a per-resolution rate is a comparison of two unrelated things.

Front deserves a correction we owe our own earlier write-ups. Its headline of "Starting at $0.05 /conversation" is the price of triage, not of an answer. Read on 22 August 2026, Front's Autopilot rate card is three tiers: $0.05 to understand and route an inquiry, $0.39 to gather context or draft a reply before handing off to a human, and $0.89 to fully resolve the conversation "without any replies from your team", with each conversation charged once at the highest outcome it reaches. That is a well-designed model and Front publishes it openly. It is simply not a five-cent product: across a realistic mix of outcomes it costs more per solved question than Fin or Help Scout, not twenty times less.

Gorgias sits in a category of its own and deserves the flag. Its pricing page states that each automated interaction also counts as one helpdesk ticket, so a resolution consumes the AI fee and a ticket from your allowance at the same time. The full picture is on our Gorgias pricing breakdown. Tidio's structure is unusual in a different way, and the plan tiles are only a third of it, which is why we pulled the whole thing apart on the Tidio pricing page.

Does Tidio charge per conversation or per resolution?

Per conversation on every tier except Premium. Lyro is billed when it engages a customer, whether or not it answers the question. Premium is the only Tidio plan that lists pay-per-resolution billing, and it also carries a guaranteed 50 percent Lyro resolution rate. Premium starts from 3,000 AI conversations and requires contacting sales.

That reservation is worth pausing on. Tidio is confident enough in Lyro to guarantee a resolution floor and to bill on outcomes, but only for customers large enough to negotiate. Everyone below that line pays for attempts. If you are evaluating the Starter or Growth tier, budget on conversations Lyro will pick up, not on problems you expect it to close.

How do you compare a per-conversation rate to a per-resolution rate?

Divide the per-conversation rate by your expected resolution rate. If a vendor charges $0.20 per conversation and the AI resolves half of them, each problem it actually solved cost you $0.40, because you also paid for the one it did not solve. Per-resolution vendors need no adjustment, since the rate already is the cost per solved question.

Run that on the published numbers and the ranking changes. Here is every per-attempt vendor normalized to cost per solved question, at a conservative 50 percent and at the 67 percent Tidio advertises for Lyro.

Two things fall out of that table. Tidio's website rate is not the bargain the sticker suggests: at its own advertised resolution rate it costs within two cents of Fin per solved question, and at a realistic 50 percent it is more expensive than Gorgias Pro. And Tidio's Shopify listing is genuinely cheap, which raises the awkward question of why the same AI agent costs more than three times as much per conversation depending on which storefront you buy it through.

Why does the resolution rate matter so much?

Because it is the only variable in the comparison you cannot see before you buy. Vendors publish rates, not their performance on your content. A per-resolution contract puts that risk on the vendor. A per-conversation contract puts it on you, which is exactly why the cheaper-looking unit is usually the per-conversation one.

This is also where the quality of your source material stops being an abstract concern and becomes a line item. An AI agent trained on a thin, out-of-date help center will engage every conversation and resolve few of them, and under per-conversation billing you pay full price for that. Getting your existing documentation into clean, structured text the agent can learn from is a real prerequisite; if it lives across a sprawling site rather than in a tidy knowledge base, a scraping API that returns LLM-ready data will do that job faster than anyone copying pages by hand. Under per-resolution billing a weak knowledge base costs you nothing but a poor customer experience. Under per-conversation billing it costs you cash every month.

Is Tidio cheaper than Intercom?

On the plan itself, comfortably. Tidio's Starter is 24.17 a month with ten seats included, while Fin carries a $29 per seat monthly minimum before a single resolution is billed. On the AI, the answer depends entirely on how well Lyro performs on your content, and at Lyro's own advertised 67 percent the two land within two cents of each other per solved question.

The seat count is the part most comparisons miss. Ten seats on Tidio's Free, Starter and Growth tiers is genuinely generous next to Zendesk at $55 to $115 per agent or Front at $25 to $105 per seat. If your costs are driven by headcount, Tidio wins easily. If they are driven by AI volume, the meter catches up with you. Our Intercom alternative comparison works through the Fin side of that in detail.

What does a Lyro conversation that ends in a handoff cost?

The same as one Lyro solves. Tidio's meter counts conversations the AI engages, not outcomes, so a handoff to a human consumes an AI conversation and then a billable conversation on the human meter as well. Ask a Tidio rep to confirm this against your contract, because it is the single number that decides whether the plan is cheap.

Contrast the language Intercom uses on the same scenario. Fin's pricing page states plainly that you are not charged when a conversation is simply passed to your team without an outcome. Gorgias goes further in the other direction: if a human agent replies within 72 hours of an AI resolution, Gorgias charges that interaction only as a billable helpdesk ticket rather than as an AI interaction. Three vendors, three different answers to the same question, and none of them is guessable from the price.

Which AI billing model is cheapest overall?

Per-conversation billing is cheapest when your AI resolution rate is high and your volume is modest. Per-resolution billing is cheapest when the AI is unproven on your content. Flat subscription is cheapest at volume, because it is the only model where getting better at deflection lowers your cost per contact instead of raising the invoice.

That last point is the one worth sitting with. Every metered model, whether the unit is a conversation, a session or a resolution, shares a structural feature: your bill goes up as the AI gets better at its job. A support team that improves its knowledge base, tightens its macros and lifts deflection from 30 to 60 percent has just doubled its AI line item under per-resolution billing. We took the opposite side of that trade deliberately, and the reasoning is in our comparison of chatbot pricing models.

What should you ask before signing?

• What is the unit, in writing? Conversation, session, interaction and resolution are four different things and vendors use them loosely in marketing and precisely in contracts.
• What happens on a handoff? Get the answer for a conversation the AI opens and a human finishes. That is the most common outcome in month one.
• Is there a minimum? Fin applies a 50-outcome monthly minimum for non-Intercom help desks. Minimums turn a usage-based quote into a floor.
• Which rate is on the order form? Gorgias publishes $0.90, $1.00 and $1.50 on the same page without reconciling them. Ask which one your contract carries.
• Does the AI unit also consume a ticket? On Gorgias it does. That doubles the effective rate and it is stated only in the fine print.
• What is the trial actually long enough to prove? Tidio and Gorgias both run 7 days. Fin runs 14 with unlimited outcomes and no card. Load real historical tickets on day one either way.

The short version

Tidio Lyro is priced attractively and metered aggressively. Below the Premium tier you pay for AI attempts rather than AI outcomes, which means the published rate understates the true cost per solved question by whatever fraction of conversations Lyro cannot close. At Tidio's own advertised performance that adjustment puts it level with Fin, a vendor whose sticker price is 50 percent higher. Neither is wrong to price the way it does. What would be wrong is comparing them on the sticker.

If you would rather not model any of this, the alternative is a plan with no AI meter at all. Our pricing is three flat tiers with carrier and Meta message fees passed through at cost, and the AI answering, qualifying and booking on every messaging channel is included rather than billed by the unit.