Free tool

AI coding cost calculator

Estimate what your team spends each month on AI coding tools that bill by the token, with any seat subscriptions added. Change the team size, usage and model to see how the total moves, and what a per-person monthly budget would change.

Runs in your browser. Nothing you type or paste leaves this page.

Team

Developers

Working days a month

Usage per person

Start from a profile

LightTypicalHeavy

The profiles are rough assumptions, not measurements. Replace them with numbers from your own tools’ usage reports where you can.

Sessions a day

Input tokens per session

Everything sent to the model across the session, including context re-sent on every turn.

Output tokens per session

Input read from the cache

The share of input tokens billed at the cached-input price.

Model

Model and price

Per million tokens: input $2.00, cached input $0.10, output $10.00.

Seats, spread and budgets

Seat subscription per person a month

USD

Your own plan price, if you pay per seat. Leave at 0 for API use only.

The heaviest 10% use this many times the average

1 means everyone uses the same. The total stays the same; only the spread changes.

Cap each person’s monthly API spend

Monthly estimate

$541.80

a month for 10 people, $54.18 each

API usage: $541.80
Seats: $0.00
API cost per session: $0.43
Likely range: $406.35 to $812.70

The range assumes usage 25% below to 50% above your numbers; seats don’t change.

Spread of API spend

Heaviest 1 person: $162.54 each
Everyone else (9): $42.14 each

Prices from the providers’ pages, checked on 22 October 2026: Anthropic pricing, OpenAI pricing, Google pricing. Standard rates in US dollars, before tax and any discounts. Check the current prices before you rely on an estimate.

How the estimate works

Cost of a session = input tokens × (1 − cached share) × input price + input tokens × cached share × cached-input price + output tokens × output price, with prices per million tokens.

API cost per person = cost of a session × sessions a day × working days. The team’s API cost is that times the number of people, and the seat subscription is added per person.

The range runs from 25% below to 50% above your usage, because usage moves from month to month. Seats stay fixed.

Uneven usage. The heaviest 10% of people (at least one, rounded up) use the multiple you choose of the average, and everyone else shares the rest equally, so the total doesn’t change. A budget cap then holds each person to the limit: the estimate counts how many people go over it and how much spending stops there.

What it leaves out. Cache-write charges, long-prompt price tiers, batch and regional pricing, taxes and discounts. For when prompt caching pays off, see the prompt cache break-even calculator.

If you run SpanDock, its model proxy can hold each person to a budget for a day, a week or a month: near the limit only cheaper models are used, and over it calls are blocked, sent to the cheapest model or just flagged. Per-person budgets come with the Standard plan and above.