Let’s build your chatbot today!
Launch a no-code WotNot agent and reclaim your hours.
*Takes you to quick 2-step signup.
The cost of building a chatbot depends on one decision you haven't made yet.
Are you planning to use a chatbot platform, hire someone to build it for you, or develop it from scratch?
It’s pertinent to know the direction because each path comes with a completely different price tag.
A no-code platform might cost less than $100 a month, while a custom-built chatbot can require tens of thousands of dollars upfront. I can’t inherently point to a singular option as better because they’re built for different needs.
That’s why there isn’t a single answer to the question, “How much does it cost to build a chatbot?” The right answer depends on what you’re trying to build, how much control you need, and who will be responsible for maintaining it.
In this guide, I'll break down the three most common ways to get a chatbot, what each one actually costs, the ongoing expenses you should expect after launch, and when each approach makes the most sense.
Three Ways to Get a Chatbot (and What Each Actually Costs)
Before we get into the details, here's the quick version. What a chatbot costs depends almost entirely on which of these three paths you take:
Path | Upfront Cost | Monthly Cost | Time to Launch | Best For |
No-code platform (build it yourself) | $0 | $29 to $500/month | Days to weeks | Teams that can manage their own bot with a visual builder |
Platform + managed services | $12,000/year | Included in annual plan | 2 to 4 weeks | Teams that want a production bot without building it themselves |
Custom development | $15,000 to $75,000+ | $500 to $2,000/month ongoing | 2 to 6 months | Enterprises with unique requirements and engineering resources |
These aren't made-up ranges. They're based on real platform pricing, real agency quotes, and the actual cost of running LLM APIs at production volume. I'll break each one down below.
Let’s build your chatbot today!
Launch a no-code WotNot agent and reclaim your hours.
Let’s build your chatbot today!
Launch a no-code WotNot agent and reclaim your hours.
Path 1: Use a No-Code Platform
This is the most common starting point, and for most businesses, it's the right one.
You sign up for a chatbot platform, build your bot using a visual drag-and-drop builder, connect your knowledge base or help docs, and deploy it to your website or WhatsApp. The platform handles the AI, the hosting, the infrastructure. You handle the conversation design and the content.
The monthly cost typically falls between $29 and $500 depending on conversation volume, AI credits, and the features you need.
To give you a real example, WotNot's pricing breaks down like this:
Lite ($29/month): 1,000 chats, 1,000 AI credits, no-code builder, Zapier integration. Enough for a small business testing the waters.
Starter ($99/month): 5,000 chats, access to all LLM models, AI Studio, integrations, ticketing support. This is where most growing teams start.
Premium ($299/month): 10,000 chats, live chat, white-label widget, HTTP requests. For businesses deploying across multiple customer touchpoints.
Other platforms in this range include Tidio, Chatbase, and Botpress (which starts free but jumps to $189/month once you need real conversation volume).
What's included that you'd otherwise pay for separately: LLM inference, knowledge base hosting, the chat widget, deployment to channels. On most modern platforms, AI usage is bundled into the subscription rather than billed per token. That's a significant shift from even a year ago, when usage-based AI fees were the biggest source of surprise bills.
What you give up: deep customization over the AI pipeline, self-hosting options, and the ability to fine-tune retrieval logic or swap embedding models. For most support, lead gen, and FAQ use cases, you won't need any of that.
Time to launch: A basic bot can be live in a day. A well-configured bot with proper knowledge base training and testing takes one to two weeks.
Path 2: Use a Platform With Managed Services
Most people think managed services simply mean "someone builds the chatbot for you."
That's true. But it misses the bigger idea.
The real value of managed services is that you're buying expertise, not just software.
Here's what many businesses discover after signing up for a chatbot platform. The software gives you everything you need to build an AI chatbot, but it assumes you'll figure out the rest yourself.
Things like:
How should the conversation be designed?
What should the AI answer, and what should it avoid?
Which integrations actually matter?
How do you test whether the chatbot is performing well?
Why is it giving the wrong answer to certain questions?
How do you improve it after launch?
Those aren't software problems. They're implementation problems.
Most businesses simply don't have someone who knows how to solve them.
Think about an HR team that wants to automate employee queries. They know their policies inside out, but they probably don't know how to design an AI agent, structure a knowledge base for retrieval, or measure whether the chatbot is actually helping employees. The same is true for support, marketing, and operations teams.
That's where managed services come in.
You're not paying someone because the platform is difficult to use. You're paying for the experience of people who have built, tested, and optimized chatbots before. Instead of becoming chatbot experts yourself, you get a team that handles the implementation while your team focuses on the business.
When managed services make sense
This approach is a good fit if:
Your team doesn't have anyone with chatbot or AI expertise.
The project keeps getting delayed because nobody has time to build it.
The chatbot will support business-critical workflows like customer support, lead qualification, or appointment booking.
You need integrations with tools like your CRM, help desk, or booking system.
You want ongoing optimization instead of launching the chatbot and hoping it performs well.
A managed service provider typically handles planning, implementation, testing, deployment, and continuous optimization.
For example, WotNot's managed plan starts at $12,000 per year. It includes the platform subscription, 20,000 chats, a dedicated account manager, 60 hours of managed services, and quarterly performance reviews. That works out to roughly $1,000 per month for a production-ready chatbot that's professionally implemented and continuously supported.
Other chatbot providers and consultancies offer similar services, typically ranging from $1,500 to $3,000 per month, depending on the complexity of the implementation and the level of ongoing support.
When managed services don't make sense
If you're building a simple FAQ chatbot or a basic lead capture bot that your team can set up in a few days, managed services are probably unnecessary. A no-code platform will usually get you there faster and at a much lower cost.
At the other end of the spectrum, if you need complete control over the AI architecture, custom retrieval logic, self-hosting, or highly specialized workflows, a fully custom development project is likely the better choice.
Path 3: Build a Chatbot From Scratch
Building a chatbot from scratch gives you something that a platform can’t give, which is complete ownership.
But ownership comes with responsibility.
When you choose the custom route, you're not just building a chatbot. You're building and maintaining an entire AI ecosystem. Your team becomes responsible for everything from the AI model and retrieval pipeline to the infrastructure, integrations, security, monitoring, and future updates.
That's why custom development is usually reserved for businesses with requirements that existing platforms simply can't meet.
What you're actually building
A custom chatbot typically includes:
The chat interface
LLM integration
RAG pipeline and vector database
Backend services and APIs
Authentication and user management
CRM, ERP, or internal system integrations
Hosting and deployment infrastructure
Monitoring and analytics
Every one of these components needs to be designed, built, tested, and maintained.
What does it cost?
The final price depends on how much of that stack you're building.
Project | Typical Cost |
Rule-based chatbot | $5,000–$15,000 |
AI chatbot with a knowledge base | $20,000–$80,000 |
Enterprise AI assistant | $80,000–$250,000+ |
These are upfront development costs.
The ongoing costs begin after launch.
A note on who's quoting. Nearly every source for these figures is a development agency, and they have an obvious interest in quoting high. Kellton, for instance, puts basic rule-based bots at $30,000–$80,000, far above everyone else. The ranges above lean on the median across sources rather than any single quote.
A note on geography. Rates vary enormously by where your team is. Clutch's 2026 data shows chatbot developers charging anywhere from under $25 to around $200 an hour, with $50–$99 the most common band. The same build can differ by 3x depending on whether you hire in North America, Eastern Europe, or South Asia.
The costs that continue every month
Unlike a chatbot platform, custom development doesn't include hosting, AI usage, or maintenance.
You'll typically pay for:
LLM API usage, which scales with conversation volume.
Vector database hosting for your knowledge base.
Cloud infrastructure to run your application.
Maintenance and updates as models, integrations, and business requirements evolve.
For example, a chatbot handling around 1,000 conversations per month may spend $50–$300/month on LLM API usage alone. At 10,000 conversations, that can increase to $400–$2,000/month, depending on the model you choose and how much context each conversation uses.
You'll also need to budget for infrastructure and maintenance. Server hosting typically adds another $100–$500/month, while vector database costs can range from free to around $500/month. Most teams also set aside 15–25% of the original development cost each year for maintenance, bug fixes, security updates, and ongoing improvements.
When custom development makes sense
Building from scratch is usually the right choice if:
You need complete control over the AI architecture.
Your organization has strict compliance or self-hosting requirements.
You're integrating with proprietary internal systems.
You're building capabilities that existing chatbot platforms don't support.
You already have an engineering team that can own the product long term.
When it doesn't
If your goal is to answer questions from a knowledge base, qualify leads, book appointments, or automate customer support, custom development is often unnecessary.
Modern chatbot platforms can deliver those use cases much faster and at a fraction of the cost.
Custom development makes sense when your requirements are unique—not simply because you want an AI chatbot.
What Actually Drives the Cost Up
Once you’re thinking of how much it’d cost if you’re going to build your chatbot, the next question is what actually drives the cost.
The answer isn't everything. Some decisions can double your budget, while others sound expensive but barely move the needle.
Things that genuinely increase cost
1. Use case complexity
The biggest cost driver isn't whether your chatbot uses AI. It's what you expect it to do.
A chatbot that answers FAQs from your help center is relatively simple to build. One that needs to authenticate users, update CRM records, trigger workflows, access internal databases, and make decisions across multiple systems is a much larger project.
That single decision often determines whether a no-code platform is enough or whether you need a custom build.
2. Integrations
Connecting your chatbot to a CRM is usually straightforward. Connecting it to a CRM, a help desk, a payment gateway, and multiple internal systems is a different story.
Each integration adds development time on the custom path and can push you into higher pricing tiers on the platform path.
A good rule of thumb is to launch with only the integrations you actually need. Everything else can come later.
3. Conversation volume
More conversations mean more AI usage, which means higher costs.
On most chatbot platforms, this usually means moving to a higher pricing tier. With a custom build, it means higher API bills because you're paying the AI provider directly. Either way, conversation volume is one of the biggest ongoing cost drivers.
4. Multilingual support
Supporting multiple languages isn't just about translating responses. Every language adds more content to maintain, more testing, and more quality assurance to ensure the chatbot responds accurately.
Supporting two languages is relatively simple. Supporting fifteen is an entirely different project.
5. Compliance and security requirements
Requirements like HIPAA, GDPR, SOC 2, data residency, or on-premise deployment can significantly increase costs.
They often limit your platform options or require additional engineering work in a custom build. In many cases, compliance requirements are the reason businesses choose custom development over a chatbot platform.
Where businesses overspend
This is where many chatbot projects spend more than they need to. Not because the technology is expensive, but because the scope grows far beyond what's actually needed.
1. Don't build for every scenario on day one
One of the most expensive mistakes is trying to build the perfect chatbot before you've spoken to a single real user.
You need to start with the five or ten questions your customers ask most often. Launch it, learn from real conversations, and then expand.
The chatbot you need six months from now will almost certainly look different from the one you imagined on day one.
2. Avoid paying for features you won't use yet
Advanced analytics, sentiment analysis, multi-agent orchestration, and other enterprise features all have their place.
But if they aren't solving an immediate business problem, they can wait.
Every feature you add to the first version increases both cost and development time. Start with the features you know you'll use, then expand as your chatbot matures.
3. Don't obsess over a custom chat widget
A pixel-perfect chat widget might look impressive during the planning phase, but most users don't remember the interface.
They remember whether the chatbot solved their problem.
A standard widget with your branding is more than enough for most launches. Invest in the experience first. You can always refine the design later.
4. Don't pay frontier model prices for every message
If you're on the custom path, this is the single biggest cost lever most teams miss.
Not every message in a conversation needs your most expensive AI model. A greeting, a simple FAQ, a clarification question: these can be handled by a smaller, cheaper model. The complex turns where the bot needs to reason through ambiguous questions or pull from multiple knowledge base sections: that's where a frontier model earns its cost.
This is called model routing, and combined with caching (storing answers to frequently asked questions instead of regenerating them every time), it can cut your LLM API bill by 60 to 70% without any drop in answer quality.
If you're working with an agency or vendor on a custom build, ask them specifically how they handle model routing and response caching. If the answer is vague, they're probably running every message through the same expensive model. That's the easiest money to save.
Start building, not just reading
Build AI chatbots and agents with WotNot and see how easily they work in real conversations.
Start building, not just reading
Build AI chatbots and agents with WotNot and see how easily they work in real conversations.
Start building, not just reading
Build AI chatbots and agents with WotNot and see how easily they work in real conversations.
The Cost Nobody Talks About: After Launch
Building a chatbot is only part of the investment.
Once it's live, you'll have ongoing costs to keep it accurate, secure, and performing well. Some of these are bundled into chatbot platforms. Others become your responsibility if you build the chatbot yourself.
For custom-built chatbots, maintenance alone typically costs 15% to 20% of the original development cost each year. A chatbot that costs $50,000 to build may require another $7,500 to $10,000 annually for security updates, bug fixes, infrastructure maintenance, and keeping integrations working as connected systems evolve.
And that's just maintenance.
Keeping your knowledge up to date
Your chatbot is only as accurate as the information it can access.
As your products, pricing, documentation, and policies change, your knowledge base needs to change with them. Otherwise, the chatbot starts giving outdated or incomplete answers.
On a chatbot platform, updating your knowledge base is usually as simple as uploading new documents or syncing your content. With a custom build, your team is responsible for updating the content, regenerating embeddings where needed, and ensuring the retrieval pipeline reflects the latest information.
Paying for AI usage
Every conversation your chatbot handles consumes AI resources.
On most chatbot platforms, these costs are included or bundled into your subscription. With a custom build, you're billed directly by the AI model provider, so your monthly costs increase as conversation volume grows.
The good news is that modern models have become significantly more affordable. The right choice depends less on price and more on the complexity of the conversations your chatbot needs to handle.
Improving the chatbot over time
Launching a chatbot isn't the finish line.
Once people start using it, you'll discover questions you didn't anticipate, gaps in your knowledge base, and conversations where the AI doesn't respond the way you expected.
Someone has to review those conversations, identify recurring issues, update the knowledge base, and continuously improve the chatbot. On a managed plan, this is handled for you. On a self-serve platform or custom build, it's your team's responsibility.
For custom builds specifically, budget realistically for this. Active maintenance on a production chatbot typically runs 10 to 20 hours per month of someone's time, which translates to roughly $1,500 to $5,000/month depending on whether that's an in-house engineer or a contractor. This is the line item most vendor quotes leave out entirely, and it's the reason chatbots that worked perfectly at launch start giving wrong answers six months later.
Compliance and security
If your chatbot handles sensitive customer or employee data, compliance doesn't end after deployment.
Standards like HIPAA, GDPR, or SOC 2 often require ongoing audits, security reviews, and periodic recertification. Depending on your industry and compliance requirements, this can add $2,000 to $10,000 per year to your operating costs.
Typical Monthly Ongoing Costs
Cost Category | No-Code Platform | Managed Services | Custom Build |
Platform or hosting | Included in subscription | Included in plan | $100–$500/month |
LLM API usage | Included or bundled | Included or bundled | $50–$2,000+/month |
Vector database | Included | Included | $0–$500/month |
Knowledge base updates | Your team | Managed for you | Your team |
Monitoring and optimization | Your team | Managed for you | Your team or contractor |
Maintenance and bug fixes | Handled by the platform | Handled by the provider | 15–20% of development cost per year |
The upfront price tells you what it costs to launch a chatbot.
The total cost of ownership tells you what it costs to run one.
Before choosing between a chatbot platform, managed services, or a custom build, compare the total cost over the first year—not just the initial price tag. That's the number that will have the biggest impact on your budget.
The Real Cost Is Getting It Wrong the First Time
The most expensive chatbot isn't the one with the highest price tag. It's the one that gets built twice.
I've seen teams spend $40,000 on a custom build only to realize six months in that a $99/month platform would have handled the same job. I've also seen teams outgrow a free-tier chatbot within weeks because they didn't think through conversation volume or integrations upfront.
The cost question isn't really about money. It's about matching the approach to what you actually need. Start too small and you'll rebuild. Start too big and you'll pay for complexity you never use.
If you want a number that cuts through the noise, do the three-year math. It's where the real cost difference between paths shows up.
Path | Year 1 | Year 2 | Year 3 | Three-Year Total |
No-code platform ($99/month) | $1,188 | $1,188 | $1,188 | $3,564 |
Managed services ($12,000/year) | $12,000 | $12,000 | $12,000 | $36,000 |
Custom build ($30K + $2K/month ongoing) | $54,000 | $24,000 | $24,000 | $102,000 |
Those numbers aren't meant to scare you away from custom development. For the right use case, $102,000 over three years is a bargain compared to the cost of the problem it solves. But if a $99/month platform handles the same job, that's a $98,000 difference you could have spent elsewhere.
If you're still figuring out which path makes sense for your situation, talk to our team. We'll help you scope it honestly, even if the answer turns out to be that you don't need us.
FAQs
FAQs
FAQs
How much does a simple chatbot cost?
Is it cheaper to build a chatbot or use a platform?
What are the hidden costs of running a chatbot?
Can I build a chatbot for free?
How long does it take to build a chatbot?
ABOUT AUTHOR
Hardik Makadia
Co-founder & CEO, WotNot
Hardik leads the company with a focus on sales, innovation, and customer-centric solutions. Passionate about problem-solving, he drives business growth by delivering impactful and scalable solutions for clients.
Start building your chatbots today!
Curious to know how WotNot can help you? Let’s talk.
Start building your chatbots today!
Curious to know how WotNot can help you? Let’s talk.
