AI gross margin calculator

Model margin with both model spend and non-model COGS so finance and product teams can test whether pricing still holds under real workload assumptions.

Revenue can look healthy while model spend, infra, and review cost quietly compress gross margin below target.

Calculator inputs

Input tokens / requestAverage prompt + context tokens per request.Output tokens / requestAverage completion tokens returned per request.Requests / monthCurrent or planned monthly request volume.Input cost / 1M tokens (USD)Published or negotiated input token rate.Output cost / 1M tokens (USD)Published or negotiated output token rate.Non-model COGS / month (USD)Human review, infra, and support cost outside model spend.Monthly feature revenue (USD)Revenue attached to the feature or workflow.Target gross margin %Margin threshold the feature or workflow should hold.

Result summary

Total AI COGS

$458.72

Total monthly cost

$198.72

Non-model COGS

$260.00

Cost per request

$0.0008

Gross margin

69.4%

Margin gap vs target

-5.6%

Total AI COGS is $458.72, projected gross margin is 69.4%, and you are -5.6% against the target margin.

Why this matters operationally

Finance alignment

Use one margin model shared across product, finance, and platform teams.

COGS visibility

See how much margin pressure comes from model spend versus non-model costs.

Pricing decisions

Compare current margin against target before packaging or usage limits need to change.

How to use this estimate

Estimate model workload spend

Set request and token assumptions with current provider pricing.

Add non-model COGS

Include review, support, and infra costs that still count against feature gross margin.

Compare against target

Use projected margin and gap-to-target to decide if pricing or limits need revision.

Turn estimates into live guardrails

Move from estimated margin to tenant and feature-level margin visibility once the workload is live.

Track margin live Open tenant cost tracking