👤

AI Agent Per-User Cost Calculator

Estimate monthly AI agent cost per active user across LLM API, vector memory, tool calls, human review, and hosting. Compare cost of revenue, gross margin, and required seat price.

Scenario presets

Per-user usage profile

Monthly active users (MAU) Sessions per user / month Requests per session Input tokens / request Output tokens / request Cached input (%) LLM provider Storage per user (GB) Embedding tokens / user / month Embedding model Vector store Tool preset Human review rate (%) Review hourly cost ($) Minutes to review one session Setup hours Setup hourly rate ($) Maintenance hours / month Price per user / month ($) Projection (months)

Estimates are directional. Last updated: 2026-07-08. See notes.

Cost per user / mo

—

Cost per session

—

Gross margin

—

Total monthly cost

—

LLM API cost / user / mo

—

Memory & storage / user / mo

—

Tool calls / user / mo

—

Human review / user / mo

—

Setup cost amortized / mo

—

Required price for 70% margin

—

Monthly cost breakdown

Cost line	Per user / mo	Total / mo	Share
LLM API tokens	—	—	—
Embedding + vector storage	—	—	—
Tool calls	—	—	—
Human review	—	—	—
Maintenance	—	—	—
Variable cost per user / mo	—	—	—

Verdict

—

Frequently asked questions

Why calculate cost per user instead of per request?

Per-user economics make it easier to set subscription pricing, compare against customer lifetime value, and understand whether a fixed-price seat covers variable AI spend.

What is included in the monthly per-user cost?

LLM API tokens (input, cached input, output), vector/embedding storage and updates, tool calls, human review labor, and ongoing maintenance. One-time setup is shown separately and amortized over the projection period.

How is gross margin calculated?

Gross margin = (price per user per month - variable cost per user per month) / price per user per month. It excludes setup and fixed overhead.

Does this include local GPU or cloud hosting cost?

It includes vector storage hosting and tool-call rates. For full hosting infrastructure sizing, pair this with the Agent Hosting Cost Comparator or Inference Capacity Planner.

What if my product is free or usage-based?

Set price per user to $0 or use the results to decide what price or conversion rate you need to break even. The cost-per-session metric helps with usage-based billing.

API Pricing Comparator → Memory / RAG Cost Calc → Tool Call Cost Calc → Startup Runway Calc →

Per-user cost estimates are directional. Replace provider rates and usage assumptions with your actual production data for a precise forecast.