/

AI & ML

AI Chatbot Development Cost and Process: A Buyer’s Guide

Rohan Roy

·

AI Chatbot Development Cost and Process: A Buyer’s Guide | Sterling Technolabs blog cover

A chatbot quote can be $5,000 or $300,000, and both can be fair. The word covers a rule-based FAQ bot, an LLM assistant trained on your documents and an AI agent that processes refunds inside your systems. Your AI chatbot development cost comes down to three choices: which of those you need, how much of your own data it must use, and which channels it runs on.

This buyer’s guide breaks down each type, the build process from discovery to launch, what each cost tier includes, the running costs, and the metrics that tell you it’s working.

Rule-Based Bot, LLM Chatbot or AI Agent?

There are three main types. The choice between them, often called AI agents vs chatbots, shapes your whole plan.

Factor

Rule-based bot

LLM chatbot

AI agent

How it works

Follows fixed scripts and menus

Uses a language model to answer from your content

Uses a model to plan steps and use tools

What it can do

Answer set questions, collect basic details

Answer free-form questions in plain language

Check orders, book slots, update records

Data needed

Scripted flows

Help docs, FAQs, policies

Your content plus live access to your systems

Typical risk

Fails on any question outside the script

Gives a wrong or made-up answer

Takes a wrong action in a live system

Best fit

Simple FAQs, lead forms

Support and product questions

Tasks that need action, not only answers

Why the type sets the price

Each step to the right in the table adds work. A rule-based bot needs scripts. An LLM chatbot needs a clean knowledge base and testing for wrong answers. An AI agent also needs secure links to your systems and clear rules for what it can do.

That is why your AI chatbot cost rises with the type you choose. Pick the type based on the job, not the label.

How an AI Chatbot Gets Built, Step by Step

Here is a simple look at the six steps in a typical build. Each step adds to your AI chatbot development cost, so it helps to know what happens in each one.

Discovery: You, your support or sales lead, and the project team list the top customer questions, the goals, and the systems the bot must connect to. If this step is rushed, the bot ends up solving the wrong problem.

Knowledge base: Your content owners and a developer collect help articles, policies, and product details, then clean them up. Old or conflicting content leads to wrong answers.

Retrieval: Developers set up the search layer that finds the right content for each question before the model replies. A weak setup means the bot pulls the wrong page or misses the right one.

Channel integration: Developers connect the bot to your website, WhatsApp, Slack, or app, and to tools like your CRM. Each channel has its own limits, which can break a feature you planned.

Testing: Your team and the developers run real questions, odd cases, and security checks. Skip this, and your customers will find the errors for you.

Launch: The bot goes live, often for a small group first. Without monitoring, problems can go unseen for weeks.

Where to focus your time

Discovery and the knowledge base shape most of the result. That’s why you should spend more time on them than on the launch itself.

AI Chatbot Development Cost by Tier

Now, coming to the cost of a build, your AI chatbot development cost mostly depends on the tier you pick. The ranges below are market estimates as of October 2026. Your final quote will change with your scope, data, and integrations.

Tier

Scope

Build cost

Monthly run cost

Timeline

Basic

Rule-based bot for FAQs and lead forms

$5,000 to $15,000

$100 to $500

3 to 6 weeks

Mid

LLM chatbot trained on your content, 1 or 2 channels

$25,000 to $90,000

$500 to $2,000

8 to 12 weeks

Advanced

AI agent with system access, several channels, and compliance needs

$90,000 to $300,000+

$2,000 to $10,000+

12 to 20 weeks

What this looks like at Sterling Technolabs’ rates: at our 2026 US rates (mid-level developers $20 an hour, senior engineers $25, testers $15), a basic bot built by one developer with part-time testing over 3 to 6 weeks comes to roughly $3,000 to $7,000 in engineering time. A mid-tier LLM chatbot with a senior and a mid-level engineer plus part-time analysis and testing over 8 to 12 weeks lands at about $21,000 to $31,000. Model and hosting costs are billed separately, and every AI project is scoped and quoted on its own, so treat these as planning figures, not a quote.

Sterling Technolabs Pro Tip: Ask vendors to split build cost from monthly run cost. A low quote with high usage fees can cost more.

What changes the price inside a tier

Integrations: Each system you connect, like a CRM or order tool, adds build time.

Data quality: Messy or outdated content needs cleanup before the bot can use it.

Compliance: Rules like GDPR or HIPAA add review and security work.

Languages and channels: More of each means more testing.

Build or buy?

A ready-made tool can be enough if your needs are simple. Many cost from about $20 to a few hundred dollars a month, and you can launch in days. A custom build makes sense when the bot must use your private data, connect to your systems, or follow strict security rules. In that case, your AI chatbot cost is higher at the start, but you get more control.

What You Pay After Launch

Once you have a live bot, your AI chatbot cost does not stop. Monthly costs usually fall into these groups:

Model usage: You pay for each conversation the AI model handles. More chats and longer answers cost more.

Hosting: Servers and databases that keep the bot running.

Monitoring: Tools and people that check answer quality and errors.

Content upkeep: Updating the knowledge base when prices, policies, or products change.

Maintenance: Fixes and updates to the bot, often planned at 15% to 20% of the build cost each year.

Example: 1,000 conversations a month

Item

Estimated monthly cost

Model usage

$20 to $100

Hosting and tools

$100 to $300

Monitoring and content upkeep

$200 to $600

Total

$320 to $1,000

This is an example for a mid-tier bot. Your numbers will change with message length, the model you choose, and channel fees.

Channels: Website, WhatsApp, Slack, and In-App

There is no single best channel, so start where your customers or team already talk to you. Each one adds setup and testing, which raises your AI chatbot development cost.

Website: The easiest to set up, since it is a chat widget on your site. It works well for support questions and lead capture.

Slack: Needs workspace permissions and a security review. It is best for internal help desk and HR questions.

In-app: Built into your mobile app or product, so it needs app developers. It suits support that happens inside the product.

WhatsApp needs a closer look

A WhatsApp AI chatbot meets customers on an app they already use. The fees changed recently. Based on Meta’s WhatsApp pricing documentation:

  • From October 1, 2026, free-form replies sent inside the 24-hour customer service window are charged per message. The rate matches the utility and authentication rate in each country.

  • Replies from Meta’s own AI, called Meta Business Agent, are billed by token use instead.

  • Each message gets one charge, never two.

  • Meta may update rates as often as every quarter.

That’s why you should keep bot replies short and clear. Fewer back-and-forth messages mean lower fees. Check the current rate for your country before you set a budget.

Training the Bot on Your Own Data

You need to give your bot the right content before it can answer well. Outdated or missing content is one of the most common reasons bots give wrong answers.

What data to use

  • Help articles and FAQs

  • Return, shipping, and billing policies

  • Product details and pricing

  • Past support tickets, to find the questions customers really ask and the answers that worked

Clean it first

Remove old pages, fix answers that conflict, and delete private details like names and card numbers. This work adds to your AI chatbot cost, but it saves you from wrong answers later.

Retrieval or fine-tuning

Most business bots use retrieval. The bot searches your content for each question, then writes an answer from what it finds. When a policy changes, you update the document and the bot uses the new version right away.

Fine-tuning retrains the model itself. It costs more and is slower to update, so it is best kept for special cases, like a fixed brand voice.

Keep it current and safe

Assign an owner to review the content on a set schedule. If you handle health, payment, or EU customer data, plan for extra security and compliance review. That review adds to your budget.

Learn more about AI chatbot for travel & health: A toolkit for tailored wellbeing

For a worked example, see our blueprint for an AI chatbot for a real estate brokerage that answers listing questions from MLS data, qualifies buyers and sellers, books showings and hands each lead to the right agent. It is an 8 to 10 week build with a team of four, which sits in the mid tier above.

Measuring Success: Containment, CSAT and Handoff

To be able to efficiently judge your bot, track a few numbers every week.

Metric

What it shows

Formula

Containment rate

Chats the bot finished without a human

(Total chats minus chats passed to a human) ÷ total chats × 100

CSAT

How happy customers are with the chat

Positive ratings ÷ total ratings × 100

Handoff rate

Chats passed to a human agent

Chats passed to a human ÷ total chats × 100

Sterling Technolabs Key Insight: A high containment rate can hide failures. Check it next to resolution and repeat contacts before calling the bot successful.

Pair containment with resolution

A customer can leave a chat frustrated without asking for a human. That chat still counts as contained. That’s why you should also track two more numbers:

Resolution rate: The share of chats where the issue was actually solved.

Repeat contacts: Customers who come back about the same issue within 7 days.

Check handoff quality

A good handoff sends the full chat to the agent, so your customer does not have to repeat anything. Track CSAT for handed-off chats on its own.

Set your own targets

Benchmarks differ by industry and by how each vendor counts. Measure your first 90 days, then set goals by topic. Topics with low resolution can also show you where the AI agents vs chatbots question matters, since those may need actions.

When to Upgrade a Chatbot to an AI Agent

If you wish to know when a chatbot is no longer enough, look at what your customers ask for. The AI agents vs chatbots choice comes down to answers versus actions. Here are the signs that it is time to upgrade:

  • Customers often ask for things the bot cannot do, like refunds, bookings, or order changes.

  • Your team keeps doing the same manual task after every chat.

  • Many chats are passed to a human only because the bot has no access to your systems.

  • Resolution is low on topics that need account or order details.

What changes when you upgrade

An agent can take actions in your systems, so a mistake can have real effects. You will need clear limits on what it can do, approval steps for sensitive actions, and a full log of each action. Testing also takes longer.

That’s why your AI chatbot development cost goes up with an agent. Start with one or two low-risk actions, like checking an order status, and add more once they work well.

Next Steps

Pick the chatbot type by the job it must do, not by the label. If your customers only need answers, start with a trained chatbot. If they need actions, plan for an agent. That’s how you keep your AI chatbot development cost in line with what your business needs.

Your first step is simple. This week, list your top 20 customer questions and mark which ones need an answer and which ones need an action.

Sterling Technolabs builds chatbots and AI agents for businesses, and you can read about our AI software development services if you want help scoping your use case and budget.

FAQs

How much does AI chatbot development cost?

AI chatbot development cost ranges from about $5,000 for a basic rule-based bot to $300,000 or more for an advanced AI agent. Most mid-tier chatbots fall between $25,000 and $90,000.

How much does it cost to run an AI chatbot each month?

Running costs fall between $100 and $10,000 a month, depending on the tier. They cover model usage, hosting, monitoring, and content updates, so your AI chatbot cost continues after launch.

How long does it take to build an AI chatbot?

A basic bot takes about 3 to 6 weeks, a mid-tier chatbot about 8 to 12 weeks, and an advanced AI agent about 12 to 20 weeks. Integrations add time.

What is the difference between a chatbot and an AI agent?

A chatbot answers questions, while an AI agent takes actions, like processing a refund or booking a slot. With AI agents vs chatbots, agents cost more and carry more risk.

How much does a WhatsApp AI chatbot cost?

A WhatsApp AI chatbot costs the build price plus Meta fees. Since October 1, 2026, free-form replies in the 24-hour window are charged per message, and rates vary by country.

Watercolour meadow landscape

Send us your brief. We’ll help turn it into reality

A senior engineer replies within one business day with questions or a first plan: approach, team, timeline and estimate.

Let’s talk

Prefer WhatsApp?

Chat with our team

WhatsApp

or leave us a short brief

No spam, no mailing list. NDA available on request.