Kayako Pricing Explained (2026)
Kayako publishes exactly one number: $1 per ticket its Kay AI agent resolves. The Kayako One platform fee is quote-only, which makes forecasting the hard part.
| Plan | Price | Billing | What you get |
|---|---|---|---|
| Kay AI Agent (standalone) | $1 per resolved ticket | pay per resolution, no seat or setup fee | AI resolution layer that can sit on top of an existing help desk. Charged only on full resolutions with no human involvement. |
| Kayako One | Custom quote | contact sales | Unified inbox, ticketing, knowledge base, live chat widget, automation rules, reporting and analytics, Kay AI agent, white-glove onboarding |
One published number, one blank
Kayako's pricing page is unusually confident about one thing and silent about everything else. Kay, the AI agent, costs $1 per ticket it fully resolves — no seat fees, no setup fees, nothing charged when it escalates or fails. That is stated plainly and repeatedly.
Kayako One, the platform underneath, has no published price. The page lists what it includes — unified inbox, ticketing, knowledge base, live chat, automation, reporting, Kay access, white-glove onboarding — and then offers a demo booking.
We checked the third-party landscape and it does not resolve the question. Independent reviews published this year contradict each other outright: one reports a per-agent platform fee around $79, another states flatly that there is no platform fee and the model is pure per-resolution. Both cannot be right, and neither is the vendor. So the tier table above says custom quote, and it will keep saying that until Kayako publishes.
Why per-resolution pricing changes your incentives
Every other help desk in this directory charges for capacity. Kayako charges for outcomes, and that inverts a familiar dynamic.
Under per-seat pricing, deflection is free money: every ticket your knowledge base absorbs is a ticket an agent you already pay for did not handle. Under per-resolution pricing, deflection has a price tag. Push 3,000 tickets a month to Kay and you have added $3,000 a month to the bill.
This is not a trick — it is the model working. You are meant to compare that $3,000 against what 3,000 human-handled tickets cost you in agent time, and the comparison usually favours the AI by a wide margin. But it does mean the traditional support-ops instinct of "deflect everything" now has a budget line attached, and someone in finance will notice the invoice going up in the month your support got better.
Modelling the variable half
You can forecast the resolution charge precisely even without the platform price. Take monthly ticket volume, apply a deflection rate you actually believe, multiply by one dollar.
| Monthly tickets | 20% resolved by Kay | 35% resolved by Kay | 50% resolved by Kay |
|---|---|---|---|
| 1,000 | $200/mo | $350/mo | $500/mo |
| 5,000 | $1,000/mo | $1,750/mo | $2,500/mo |
| 15,000 | $3,000/mo | $5,250/mo | $7,500/mo |
Be conservative on the rate. Deflection performance depends almost entirely on how good your knowledge base is and how repetitive your queue is. A password-reset-and-shipping-status queue with well-maintained docs will clear 40%. A queue full of account-specific billing disputes will not clear 15%, and no vendor's AI changes that.
Run the numbers at your realistic rate, then ask sales for the platform figure and see whether the combined total lands under what you pay now.
Where the model shines and where it doesn't
It shines for teams with high volume, high repetition and decent documentation. Ecommerce support, SaaS tier-one, anything with a long tail of the same twenty questions. Every dollar spent there replaces several dollars of agent time, and the escalation carve-out means you are not paying for the bot's failures.
It struggles for the opposite shape. Small teams with low volume get little from a resolution meter, and whatever the fixed platform cost turns out to be will dominate their bill. Complex B2B support with bespoke tickets will see low deflection and should evaluate Kayako on its inbox and ticketing merits alone, treating Kay as incidental.
The standalone option is the useful hedge here. Because Kay can run on top of an existing help desk, you can test the deflection rate against your real queue for a month, get an actual number rather than an estimate, and only then decide whether the full platform migration is worth having.
Before you sign
Get three things in writing: the platform fee at your agent count, the precise contractual definition of a full resolution, and whether there is any monthly minimum on resolutions. The second one is where the money is — the difference between "resolved without human involvement" as a technical event and as a customer-satisfaction outcome is worth real money at volume.
Kayako has changed its commercial model significantly since the Classic product, and the current page publishes less than it used to. Treat kayako.com/pricing as the authority and assume anything you read elsewhere, including here, may be behind.
What the pricing page doesn't show
- The platform price is not published anywhere on kayako.com. Third-party sources disagree sharply — some report a per-agent fee, others claim there is no platform fee at all — so the only way to know your cost is a sales call.
- A $1 resolution charge means your bill rises when the AI works well. A successful deflection campaign that moves 3,000 tickets a month to Kay adds $3,000 a month, and that is the intended outcome, not a failure.
- What counts as a full resolution is defined by Kayako, not by you. The definition — resolved without human agent involvement — is the single most important clause in the contract and deserves reading closely.
- White-glove onboarding is bundled into Kayako One rather than sold separately, which sounds generous but means its cost is inside a number you cannot see or compare.
- Kayako Classic customers migrating to the current SaaS product are moving to a different commercial model entirely; old renewal pricing is not a guide to new pricing.