Cost formula
API monthly = conversations/day × days × replies/conversation × cost(input+output tokens); Escalations = conversations × (1 − resolution rate); Total = API monthly + escalations × human handoff $; Blended = Total ÷ monthly conversations
Calculator
Enter conversations per day, replies per ticket, token sizes, and bot resolution rate to see monthly API cost, human handoff cost, and blended cost per conversation.
Support chatbot
$33,630.60/ month
$2.80 / conversation · GPT-4o mini · 65% bot resolution
Inputs
Estimate LLM spend plus residual human cost for escalated tickets.
Insights
API cost is usually small next to escalations—containment rate matters most.
Human handoff cost is your planning figure (salary + tooling + overhead), not an API price. Raising resolution rate usually saves more than switching to a cheaper model.
Models AI chatbot API spend plus optional human escalation cost for support volumes—not CRM, telephony, or seat licenses. Updated 2026-07-31. Approximate static list prices for planning only. Always verify against each provider’s official pricing page before production budgeting.
Guide
Estimate what an AI customer-support chatbot really costs: LLM replies, optional RAG context, bot containment rate, and residual human handoff cost. Most teams discover API spend is small next to escalations—so resolution rate and handoff cost dominate the blended cost per conversation.
API monthly = conversations/day × days × replies/conversation × cost(input+output tokens); Escalations = conversations × (1 − resolution rate); Total = API monthly + escalations × human handoff $; Blended = Total ÷ monthly conversations
Searchers asking “how much does an AI chatbot cost?” usually need unit economics for support, not a raw token table. This calculator turns tickets, containment, and model choice into a monthly bill and a cost-per-conversation figure finance and CX teams can share.
It depends on conversation volume, tokens per reply, model rates, and how often humans still take over. Enter your numbers here to get API + handoff totals rather than a single generic quote.
For most support volumes, human escalations dominate. That is why containment rate is the main lever once you have a capable model.
No. It models LLM API spend and an optional per-ticket human handoff cost. Add CRM and telephony separately.
CentsPerToken uses approximate list prices for planning. Confirm current provider rates before production budgeting.