How Much Does a Custom AI Chatbot Cost?

It's the first question almost everyone asks, and the honest answer is that price tracks scope: it depends on what the system actually has to do. A bot that answers ten FAQs and a bot that reads your full product catalog, resolves live order status, and escalates to a human are fundamentally different builds, and they price accordingly.

So rather than leave you with a vague "it depends," here is a precise breakdown of the real 2026 price ranges, the variables that move the number up or down, and exactly what you'll pay each month once it's live.

The three pricing tiers, at a glance

In practice, custom chatbot projects fall into one of three tiers. Hover any card to see what's included:

ranges are typical freelance pricing · your exact quote depends on scope & data

What actually drives the price

The tier you land in comes down to five variables. These are what a competent developer is actually pricing:

more of each = higher price · most cost sits in data and integrations

1. Your data

The largest hidden cost isn't the AI, it's structuring what the bot needs to know. When your policies, sizing charts, and FAQs are already documented clearly, the build is faster and cheaper. When that knowledge lives only in someone's head, it has to be captured and written down first, and that work is real.

2. Integrations

A bot that only answers questions is one price. A bot that retrieves live order status, reads your booking calendar, or writes to your CRM must be wired into each of those systems, and every integration is a distinct piece of engineering with its own cost.

3. Design & brand

A default chat bubble is essentially free. A widget that matches your brand, behaves correctly on mobile, and is genuinely pleasant to use takes deliberate design work, and it earns its keep, because a clunky interface erodes the very trust the bot exists to build.

4. How capable it needs to be

A scripted FAQ bot is cheap but brittle. A RAG assistant that answers from your real data with sources is the modern standard, and it's why most serious projects settle in the middle tier. (If RAG is new to you, I break it down plainly in this post on stopping chatbots from making things up.)

5. Support after launch

A one-off build is cheaper up front. Ongoing care, refreshing its knowledge, monitoring answer quality, and refining it as you learn what customers actually ask, is an optional monthly investment that keeps it accurate over time.

The costs nobody mentions: ongoing fees

Beyond the build itself, a chatbot carries two modest recurring costs:

• Hosting, where the bot runs. Typically $0–$20/month for a small business site.
• AI usage, the language model bills per message. For most small businesses that lands at a few dollars to a few tens of dollars a month, scaling directly with how many people actually chat.

The build is a one-time cost. Running it is almost always cheaper than a single support salary, for a bot that works every hour of every day.

How to avoid overpaying

• Start with your top 10 questions. There's no need to feed it everything on day one, launch focused, then expand deliberately.
• Don't pay enterprise prices for a small-business bot. Large agencies quote large numbers for the same RAG build a competent freelancer ships in days.
• Specify integrations up front. "Just answer FAQs" and "check live orders" are very different quotes, so decide which you genuinely need now.
• Ask what happens when it doesn't know. A bot that escalates instead of bluffing is worth far more than a cheaper one that fabricates answers.

It doesn't stop at the chatbot

A chatbot lives on a website, and even a great bot on a shaky site still loses sales. That's why I build the entire system, the AI assistant and the site or Shopify store it sits inside.

This isn't e-commerce theory for me: I designed, built, and launched my own clothing brand's website end-to-end, storefront, product pages, checkout, and the customer-experience layer. So when I say I'll handle your store and the assistant on it, that's something I've shipped for myself, not only for clients.

The bottom line

A simple FAQ bot can cost a few hundred dollars; a custom RAG assistant that answers from your real data, the tier most businesses actually want, typically runs $800–$3,000, plus modest monthly running costs. The exact figure comes down to your data, your integrations, and how capable the bot needs to be.

The reliable way to get a real number isn't a price list, it's a five-minute conversation about what you're trying to accomplish.