All business decision tools
Does usage pricing cover inference and support costs?
Model revenue per request alongside tokens, model cost and platform overhead.
Example scenario
Editable demonstration inputs, not verified industry benchmarks. Enable JavaScript to calculate your own scenario.
- Requests / month
- 1,000,000
- Average billed tokens / request
- 2,500
- Blended cost / million tokens
- $3
- Revenue / 1,000 requests
- $25
- Monthly platform & support cost
- $8,000
Monthly product contribution
Example result only, not a forecast or promise.
- Monthly usage revenue
- $25,000
- Inference cost
- $7,500
- Contribution margin
- 38%
What to do with this result
Measure input, output, cache and retry usage separately before contracting; a blended rate is only a planning approximation.
Know what is inside the model.
Planning model, not personalized financial advice. Your inputs remain in your browser.
Calculation method
Revenue = requests / 1,000 × price. Model cost = requests × tokens / 1,000,000 × blended token cost. Contribution = revenue − model cost − platform/support.
Assumptions and limitations
- Example inputs are editable scenarios, not Aryo results or independently verified industry benchmarks.
- Use one matched customer cohort and consistent periods. Recurring revenue, bookings and cash are different measures.