This calculator estimates the fully loaded cost of processing one meeting with an AI meeting assistant. It combines speech processing, language-model usage, storage, and optional fixed platform expenses, then spreads those costs across the number of meetings handled each month.
Use it when comparing vendors, setting internal chargeback rates, or deciding whether a meeting assistant should run on every call or only on selected sessions. The result separates variable cost from allocated fixed cost so you can see which part changes with usage and which part improves as volume grows.
1. Enter meeting usage
Provide the typical meeting duration and expected monthly meeting count.
2. Add variable service rates
Enter speech-processing, token, and per-meeting workflow costs using the same currency.
3. Include fixed expenses
Add monthly platform, support, or infrastructure charges that should be allocated to the workload.
4. Review unit economics
Use the main result for cost per meeting and the breakdown to distinguish variable and fixed portions.
Cost per meeting = (Minutes × Speech cost per minute) + (Tokens ÷ 1,000,000 × Cost per 1M tokens) + Storage/workflow cost + (Monthly fixed cost ÷ Monthly meetings)
All monetary inputs must use one currency. Fixed cost is allocated evenly across the entered monthly volume. The model does not include staff time unless you add it to the per-meeting workflow cost or fixed monthly cost.
The result is the estimated fully loaded operating cost for one processed meeting at the entered monthly volume.
Actual invoices may differ because vendors can apply minimum commitments, tiered pricing, cached-token discounts, or separate fees.
Given: 45 minutes, $0.006 per audio minute, 18,000 tokens at $5 per million, $0.08 workflow cost, $300 fixed monthly cost, and 1,000 meetings.
Calculation: Speech = 45 × 0.006 = $0.27. Model = 18,000 ÷ 1,000,000 × 5 = $0.09. Variable cost = 0.27 + 0.09 + 0.08 = $0.44. Fixed allocation = 300 ÷ 1,000 = $0.30.
Result: $0.44 + $0.30 = $0.74 per meeting.
At this volume, fixed expenses account for about 41% of the unit cost.
Include it only if you want a fully burdened business cost. Convert review labor into a per-meeting amount and add it to the workflow cost.
The fixed monthly expense is spread across more meetings. Variable processing cost per meeting stays unchanged unless your vendor pricing also changes.
Use a blended effective cost per million tokens based on your expected mix of full-price and discounted usage.
Yes. Run the same meeting profile with each vendor’s rates and fixed fees, then compare both unit cost and monthly total.
Implementation work, unusual overage fees, taxes, currency conversion, and quality-related reprocessing are excluded unless you add them to the inputs.