Customer-Service AI Chatbot Cost Calculator (monthly API, handoffs & savings)

How it works

A chat model re-reads the whole conversation on every turn. For the k-th customer message the input is: system prompt + all earlier messages and replies + the new message, so over a conversation of t messages input tokens = t × (system prompt + message) + (message + reply) × t(t − 1) ÷ 2, and output tokens = t × reply. The cached share applies to the system prompt part. Monthly cost = conversations × API cost per conversation + platform fees + handed-over conversations × human cost per conversation.

⚠ Estimates for information only — not financial advice.

FAQ

Why does a longer chat cost more than linearly?

Each new reply needs the whole history as input, so a 10-message chat sends the first message 10 times. Summarising old turns or limiting history keeps costs down.

What should I enter as the human cost?

Agent cost per hour ÷ conversations an agent handles per hour, plus any outsourcing fee. For example $30 an hour and 8 chats an hour is about $3.75 per chat.

Checked 10 Nov 2026. Token accounting of a multi-turn chat (each request carries the full conversation history) with the provider’s list prices from our AI price file (OpenAI official pricing page, checked 10 Nov 2026).