Chatbot pricing compared: what different platforms charge
Ask three chatbot vendors for a quote and you'll often get three completely different pricing structures, not just three different numbers. One charges a flat monthly fee. Another bills per conversation. A third wants a setup fee plus a retainer. None of them are lying to you, exactly, but comparing them side by side is like comparing a gym membership to a taxi fare: the units don't match, so the sticker price tells you almost nothing on its own.
The four pricing models you'll actually run into
Most chatbot pricing falls into a handful of shapes. Flat subscription platforms, the kind you can sign up for with a credit card, charge a fixed monthly fee for a tier that caps conversations, seats, or integrations. These are the easiest to budget for and the hardest to scale past, since the next tier up often costs noticeably more for features you may only need half of.
Per-conversation or per-message pricing is the model behind most WhatsApp-based bots, because Meta itself charges businesses per conversation above a free tier, and the platform you build on usually passes that cost through plus a margin. This model is cheap when you're small and expensive exactly when you're succeeding, since a busy month means a busy bill.
Per-seat pricing charges for each human agent who can access the chatbot's dashboard or take over conversations it hands off. It's common on platforms built around a shared inbox, and it quietly punishes a growing support team instead of a growing customer base.
Then there's the custom build: an agency or freelancer designs and deploys a bot tailored to your workflow, usually with a setup fee covering the initial build and a smaller monthly fee for hosting, monitoring, and changes. You're not paying for a seat or a conversation count here. You're paying for someone to have actually mapped your process before writing a line of logic.
Why the sticker price rarely matches the real bill
A $49-a-month chatbot plan sounds appealing until you notice the fine print capping you at 500 conversations, after which every additional conversation is billed separately. A clinic that books 40 appointments a day through WhatsApp can blow past that cap in the first ten days of the month and then spend the rest of it paying overage fees that dwarf the base subscription.
The same trap works in reverse. A flashy enterprise plan with unlimited everything is a bad deal for a five-person business that handles thirty conversations a week. You'd be buying capacity you'll never use, the chatbot equivalent of leasing a delivery van to pick up your weekly groceries.
Language support is another cost that hides until you go looking for it. A platform that handles English fluently may charge extra, or simply perform worse, in Hebrew or Russian, and testing that before signing a year-long contract is worth the extra hour it takes.
What actually belongs in a fair comparison
Strip out the marketing tier names and ask each vendor the same three questions. What's the real monthly cost at your current conversation volume, not their example volume? What happens, in dollars, the month you get 30% busier than usual? And what's included versus billed separately: human handoff, analytics, CRM integration, multilingual support?
Write the answers down next to each other in a plain table. Vendors who are confident in their pricing will answer directly. Vendors who stall or redirect you to "it depends" without offering to actually calculate it are telling you something too.
The cost nobody puts on the pricing page
The biggest number is rarely the subscription fee. It's the hours your team spends writing and updating the bot's responses, reviewing conversations it mishandled, and manually doing whatever it can't yet do on its own. A $30-a-month tool that nobody maintains ends up costing more in quiet, unbilled staff time than a $300-a-month one that was actually built around how your business runs and rarely needs fixing.
That's also where the agency-built option earns back some of its higher upfront cost. You're not just buying a chat widget, you're buying someone whose job is to notice when the bot starts failing at a new kind of question and fix it before a customer complains.
Matching the model to your actual business
A small retail shop with steady, predictable WhatsApp volume usually does fine on a flat subscription, since the ceiling is easy to estimate and rarely surprising. A business with seasonal spikes, like a clinic before holidays or a shop during sales events, should be wary of per-conversation pricing unless the rate stays reasonable at peak volume, not just on an average month. A company juggling WhatsApp, Instagram, and a website widget at once, with a support team that needs visibility across all three, often outgrows the single-platform subscription model entirely and needs something built to connect them properly.
Got a repetitive process eating up hours?
Tell us about it in a couple of lines and we will show you how automation would work on your own data.
Get in touch