“How much will an AI chatbot cost?” has no single answer, because the bill depends on how you build it and how much it’s used. What you can do is break the cost into parts and estimate each one. This guide shows you how.
Option 1: an off-the-shelf chatbot tool
Many platforms let you create a chatbot for your website without coding, usually for a monthly subscription. Pricing often depends on the number of conversations or messages, the number of seats and features such as integrations or custom branding. This is the fastest route for small businesses. Compare plans by the limits that matter to you, not just the headline price.
Option 2: build your own on an AI API
If you build a chatbot yourself, or hire a developer, the costs split into four parts:
- Model usage: what you pay the AI provider per token. This scales with traffic. See how AI API pricing works.
- Hosting and infrastructure: the server or platform that runs your chatbot, plus any database for documents or conversation logs.
- Development: the one-time effort to build, test and connect it to your website or tools.
- Maintenance: updating content, reviewing conversations, fixing issues and adapting when models or prices change.
Estimating model usage
You need four numbers:
- Conversations per day you expect.
- Messages per conversation, since every message is a separate request.
- Input tokens per request, including instructions, retrieved documents and chat history, which grows with each message.
- Output tokens per request, the length of a typical answer.
Multiply conversations by messages to get requests per day, then enter everything into our AI API Cost Calculator with your model’s current prices. Run three scenarios, low, expected and high traffic, so you know your range.
A simple budgeting worksheet
Monthly model usage (from calculator): $____
Hosting and database: $____
Chatbot platform or tools (if any): $____
Maintenance time (hours × hourly rate): $____
------------------------------------------------
Monthly running cost: $____
One-time build cost: $____
Ways to keep the budget under control
- Set spending limits and alerts in your AI provider’s dashboard.
- Limit conversation length or summarize long chats.
- Use a smaller model for simple questions.
- Answer the most common questions with a fixed FAQ before calling the model.
More ideas are in How to Cut Your AI API Costs.
Key takeaways
- Off-the-shelf tools are simpler; building your own gives control but adds development and maintenance.
- Model usage scales with traffic and conversation length.
- Estimate low, expected and high scenarios and set spending alerts.

Leave a Reply