Related: What Is a Vector Database and Why It Matters for AI
AI chatbot prices in 2026 cover a wide range. A simple plan can cost almost nothing per month. A custom support agent that connects to your systems can run into tens of thousands of dollars. This guide shows the real ranges, what makes the price go up or down, and the running costs that come after launch.
Key takeaways
- A no-code platform chatbot costs about 0 to 500 dollars per month. A custom build usually costs 5,000 to 60,000 dollars up front.
- The biggest cost drivers are how many systems it connects to, how custom the answers must be, and how much testing safety needs.
- Most buyers forget ongoing costs: model API fees, hosting, and human time to maintain and improve the bot.
- Platforms are best for fast, simple use. Custom builds win when you need control, private data, and deep integration.
- Always budget 15 to 30 percent of the build cost each year for upkeep.
The three ways to get a chatbot
There are three common paths. Each has a different price shape and a different level of control.
- No-code platforms. Tools like Intercom Fin, Tidio, or Voiceflow. You pay a monthly fee or a fee per resolved chat. Fast to set up, limited control.
- Low-code plus API. You use a framework and connect a model like Claude or GPT yourself. More flexible. Needs a developer.
- Fully custom build. A team builds the bot around your data and your systems. Most control, highest up-front cost.
If you want a simple plan first, our guide on how to add an AI chatbot to your website walks through the quick options step by step.
Real price ranges for 2026
These ranges come from common vendor plans and typical agency project quotes. Your numbers will move based on scope. Treat these as planning figures, not fixed prices.
| Type | Up-front cost | Monthly cost | Best for |
|---|---|---|---|
| Basic platform plan | 0 dollars | 0 to 100 dollars | FAQ and lead capture on a small site |
| Mid platform plan | 0 dollars | 100 to 500 dollars | Support teams, more chats, more seats |
| Custom small build | 5,000 to 15,000 dollars | 200 to 800 dollars | One workflow, one or two integrations |
| Custom mid build | 15,000 to 40,000 dollars | 500 to 2,000 dollars | Support plus private data search (RAG) |
| Custom large build | 40,000 to 60,000 dollars or more | 2,000 dollars or more | Multi-system agent, strict compliance |
Some platforms now charge per resolved conversation, often about 0.50 to 1.00 dollars each. That looks cheap at low volume. At high volume it can cost more than a custom build over a year. Always model your real chat volume before you pick a plan.
What drives the cost up or down
Two chatbots can look the same to a user but cost very differently to build. These factors explain most of the gap.
- Integrations. Reading your help docs is cheap. Connecting to a CRM, an order system, and a payment tool is not. Each live integration adds build and test time.
- Private data search. If the bot must answer from your own documents, you need a retrieval setup. We cover this in our production RAG architecture guide. It adds cost but cuts wrong answers.
- Actions, not just answers. A bot that books a meeting or issues a refund needs careful guardrails and testing. Reading is safe. Doing things is riskier and costs more.
- Languages and tone. One language is simple. Many languages and a tight brand voice add review work.
- Safety and compliance. Healthcare, finance, and legal need extra testing, logging, and human review. This can add a large share to the budget.
- Volume. More chats mean more model usage fees and more support load.
The ongoing costs people forget
The build price is not the full price. A chatbot is a living system. Here is what you keep paying after launch.
- Model API fees. You pay per word the model reads and writes. A busy support bot can cost a few hundred to a few thousand dollars per month in model fees alone.
- Hosting and tools. Servers, a vector database for search, and logging tools. Often 50 to 500 dollars per month for a mid build.
- Maintenance. Docs change. Models get updated. Bugs appear. Plan for developer time every month.
- Content upkeep. Someone must keep the bot's knowledge fresh and fix bad answers users report.
A simple rule helps here. Budget 15 to 30 percent of the build cost each year for upkeep. A 20,000 dollar build often needs 3,000 to 6,000 dollars a year to stay good.
Build vs platform: how to choose
This is the choice that decides most of your cost. Use this short test.
| Pick a platform if | Pick a custom build if |
|---|---|
| You need it live in days | You need deep control of answers |
| Your needs are FAQ and lead capture | The bot must use private or sensitive data |
| You have a small or no dev team | It must connect to many internal systems |
| Low chat volume | High volume where per-chat fees add up |
A common smart path is to start on a platform, learn what users ask, then build custom once the value is clear. This lowers risk and avoids over-building on day one.
FAQ
Can I get a useful AI chatbot for free?
Yes, for simple needs. Many platforms have free or low tiers that handle FAQs and lead capture on a small site. The limits are message caps, basic answers, and weak integration. Free works to test the idea, not to run a busy support line.
Why are custom chatbots so much more expensive?
Most of the cost is not the AI model. It is the work around it. Connecting to your systems, searching your private data safely, testing actions, and meeting compliance rules all take skilled time. The model fee is often the smallest line on the bill.
How much should I keep aside for running costs?
Plan for model fees, hosting, and maintenance every month. For a custom build, a safe rule is 15 to 30 percent of the build cost per year, plus your model usage fees on top, which scale with how many chats you handle.
Working with Apex Logic
We build AI chatbots and agents that fit your real systems and your real budget. We help you choose the path that makes sense, platform or custom, and we are honest about ongoing costs before you commit. See our AI solutions or contact us for a clear, no-pressure estimate.
References
Public pricing pages from common chatbot platforms such as Intercom, Tidio, and Voiceflow, reviewed in 2026.
Model provider pricing documentation for usage-based API fees.
Apex Logic project experience delivering custom chatbots and retrieval systems.
Comments