How much does the Zendesk chatbot cost?

Zendesk doesn't price its chatbot as a single line item. You pay for a per-agent Suite or Support plan first, then add AI on top. The conversational bot and the newer AI agent features sit in paid tiers and add-ons, and the more capable AI is charged partly on resolutions (conversations the AI handles end to end). So the real cost is your seat count multiplied by the tier price, plus the AI add-on, plus per-resolution charges as volume grows. Check Zendesk's pricing page for current figures, as they change.

Does Zendesk charge per resolution for AI?

Yes, for its more advanced AI. Zendesk's basic automation is bundled into plans, but the AI agent that resolves conversations autonomously is metered. You're billed based on the volume of conversations the AI handles. That means your AI bill rises with success: the more tickets it deflects, the more you pay. It's a fair model for Zendesk, but it does couple your cost to your volume.

Is Zendesk's AI included in the base plan?

Only the lighter features are. Basic triage, answer suggestions and bots tend to sit in the standard tiers, but the autonomous AI agent and advanced AI capabilities are a paid add-on layered on top of your per-seat subscription. Two stores on the same headline plan can pay very different amounts once AI usage and seat count are factored in.

Why does per-seat plus per-resolution pricing get expensive?

Because both halves grow. Add a support agent and your base cost steps up by a full seat. Get busier and your per-resolution AI charges climb with conversation volume. The saving you expected from AI deflection partly leaks back out as the AI handles more. Flat-rate pricing decouples the bill from both seats and volume, so growth doesn't quietly inflate the invoice.

What's the flat-rate alternative to Zendesk's chatbot pricing?

Omniops charges one monthly rate per store with every capability included: no per-seat fees and no per-resolution charges. Omni reads live order data, answers customers and takes actions like refunds, with sensitive actions queued for your approval first. The founder price is £250/month for the first 10 places, half off the standard £500 plan, with a 30-day trial. See /pricing for current details.

Is Zendesk a bad choice for ecommerce?

No. Zendesk is a mature, capable helpdesk used by large support teams, and for an enterprise with many agents and complex routing it earns its price. The question isn't whether Zendesk is good. It's whether its per-seat, per-resolution model fits a small ecommerce store whose costs you'd rather keep flat and predictable as you grow.

Zendesk Chatbot Pricing: What AI Support Costs, and the Flat-Rate Alternative

TL;DR

Quick answer: Zendesk doesn't sell a standalone "chatbot price". You pay for a per-agent plan (Suite or Support), then add AI on top, and the capable, autonomous AI is charged partly per resolution, so it scales with both your seat count and your conversation volume. There's nothing wrong with that model, but it means your bill grows as you grow. Omniops takes the opposite approach: one flat monthly rate per store, every capability included, no per-seat or per-resolution charges, with sensitive actions held for your approval. Founder price is £250/month for the first 10 places (half off the £500 plan), 30-day trial. The current Zendesk numbers live on Zendesk's own pricing page, and we won't quote figures that go stale.

If you searched "Zendesk chatbot pricing" you probably wanted a single number. The honest answer is that there isn't one, and understanding why tells you most of what you need to know before you commit.

How Zendesk actually prices AI support

Zendesk is a helpdesk first. Its pricing is built around agents (seats), not conversations, and the AI sits on top of that. Three things stack up to make your bill.

1. A per-agent plan. Everything starts with a Suite or Support plan, priced per agent per month and sold in tiers. The higher tiers add routing, reporting, automation depth and SLA features. Two agents cost twice one agent. Five cost five times. This is the foundation cost, and it steps up every time you add a person to your support team.

2. AI as a paid layer. Lighter automation, such as answer suggestions, basic bots and triage, tends to come bundled into the standard tiers. The capable part, the autonomous AI agent that resolves conversations end to end, is a separate paid add-on. So the "chatbot" most people mean when they search this term isn't included in the base price. It's layered on.

3. Per-resolution charges on the advanced AI. Zendesk's autonomous AI is metered on resolutions, meaning conversations the AI handles to completion. This is the part that surprises people. Your AI cost isn't fixed. It rises with the number of conversations the AI successfully deals with. The better the AI performs, the more resolutions it logs, the more you pay.

Put together: your monthly cost is roughly (seats × tier price) + AI add-on + (resolutions × per-resolution rate). None of those three components is fixed as your store grows.

We're deliberately not printing pound or dollar figures here, because Zendesk revises them and a stale number is worse than no number. For the current rates, go straight to Zendesk's pricing page. What matters for your decision isn't this month's exact figure. It's the shape of the model.

Why the model matters more than the headline price

A per-seat, per-resolution model has a particular characteristic: your cost is coupled to your success.

Think about what happens as a store does well. You take on more orders, so more customers message you. You hire a second person to keep up, so your seat cost steps up. Your AI handles more of those conversations, so your per-resolution charges climb. Every good thing that happens to the business nudges the support bill upward.

That's not Zendesk being unfair. Usage-based pricing is a reasonable way to charge, and for a large team it can map neatly onto value. But for a small ecommerce store, it has an awkward side effect. The whole point of adding AI was to handle more volume without your costs scaling in step. With per-resolution pricing, some of that saving leaks back out. The AI deflects the ticket, and you pay for the deflection.

This is the part the headline "starting from £X per agent" never tells you. The starting price is the floor, reached only by a single-agent store with no AI. Add the AI you actually came for, add the second seat you'll need at scale, add the resolution volume that comes with growth, and the real number is some way above the floor.

For a fuller treatment of how these pricing models compare across the major tools, our honest ecommerce AI comparison walks through Gorgias, Tidio, Intercom and Omniops side by side.

To be fair to Zendesk

Zendesk earns its reputation. It's a mature helpdesk with deep routing, strong reporting, a large app ecosystem, and a decade-plus of refinement. For a support operation running many agents, complex workflows and tight SLAs, the per-seat model is normal and the AI add-on is genuinely capable. None of this is a criticism of the product.

The question isn't whether Zendesk is good. It's whether a per-seat, per-resolution structure suits your situation: a small ecommerce store that would rather keep costs flat and predictable than watch them track every uptick in volume. If you're an enterprise with a big team, Zendesk's pricing fits the way you're shaped. If you're a two-to-ten-person store, the model can work against you exactly when things are going well.

The flat-rate alternative

Omniops prices the opposite way. One monthly rate per store. Every capability included. No per-seat fee, no per-resolution charge. Your bill is the same whether Omni handles 50 conversations or 5,000 this month, and the same whether one person or five are using it.

That single decision, flat rate instead of per-seat-plus-per-resolution, is what decouples your cost from your volume. Growth stops inflating the invoice. You can let the AI handle as much as it can without watching a meter.

Here's what that flat rate covers, and what Omni does:

Reads your live order data. Omni connects to WooCommerce, Shopify and Stripe and answers from real order records, not just a static FAQ. "Where's my order #1247?" gets a real answer.
Takes actions, not just answers. When a customer is eligible for a refund, Omni can process it. When a cart's been sitting, Omni can send the recovery message. It does the work, rather than handing every task back to a human.
Queues sensitive actions for your approval. Refunds, cancellations and anything with financial weight are held for you to approve before they go through. You stay in control of the decisions that matter, while Omni handles the back-and-forth around them.
Covers more than support. The same flat rate includes abandoned-cart recovery, inventory watching and the customer conversations across your channels. One agent, not three subscriptions.

Omni is an AI agent, not a chatbot. The difference between answering questions and taking actions is the whole story, and we cover it properly in agentic AI vs chatbot.

What the flat rate costs

One price, everything included. The standard plan is £500/month. Right now the founder price is £250/month for the first 10 places, half off, and it comes with a 30-day trial so you can see Omni working on your real store before you decide.

No per-conversation counting. No per-resolution meter. No charge for adding a second person. The full, current details are on the pricing page.

Working out which model fits you

A quick way to sense-check it. Estimate, roughly, three things about the next twelve months: how many support conversations a month you'll handle, how many people will be in your support seat, and how much of that you'd want AI to take off your plate.

If those numbers are small and flat, say a single seat, modest volume and mostly FAQ deflection, Zendesk's floor price may sit close to what you'd pay, and a mature helpdesk might be worth it.

If those numbers are growing, with more volume, a second seat coming and more of the load shifting to AI, then under a per-seat, per-resolution model each of those increases pushes your bill up. That's the case where a flat rate keeps your costs still while your store moves. The busier you get, the more the flat rate works in your favour, because the bill doesn't follow the volume.

That's the decision in one line: do you want your support cost to track your growth, or stay out of its way?

Where to go next

See the current Zendesk figures on Zendesk's pricing page. We don't reprint them here so this stays accurate.
Compare the major ecommerce AI tools honestly in our Gorgias vs Tidio vs Intercom vs Omniops guide.
Understand why an AI agent and a chatbot aren't the same thing in agentic AI vs chatbot.
See Omniops's flat rate and the founder offer on the pricing page.

We make Omniops, so of course we think flat-rate pricing suits a growing store better. But the point of this piece isn't to talk you out of Zendesk. It's to make sure you understand what you're signing up for before the per-resolution line shows up on the invoice.

Sources

Pricing model described from current vendor documentation, May 2026. Exact figures change, so always check the vendor's own page for live numbers: