Free tool

What an AI workflow costs to run

Enter the volume, the models and the review time for one AI workflow. You see the provider spend, the full monthly cost and the cost per accepted result at public list prices, updated as you type.

What you get
  • Provider spend for your routing between 2 models
  • The full monthly cost, with review time and shared operation
  • The cost per accepted result
  • With a work email, every model compared and a budget for the gateway

Your workflow

The starting values are the worked example in our guide to AI cost management: 1,000 supplier assessments a month. Replace them with your own workflow and the result updates as you type.

01The workflow

One unit of work a business owner signs off, such as a reply or an assessment.

The prompt, documents and context sent with each task.

The answer the model writes back.

02Routing

The share that fails on the routine route or is sent there by rule.

Prices are dollars per million tokens, input then output, from public pricing pages on 7 September 2026.

03Review and shared costs

The share that passes review.

Checking and correcting, sampled rather than self-reported.

Fully loaded.

This workflow's share of the gateway, hosting and monitoring.

Per result$1.47See result

Cost per accepted result$1.47

of which $0.02 is provider usage

  • Provider usage$15.74
  • Review time$800
  • Shared operation$360
Total a month
$1,176
Accepted results a month
800
All tasks on the routine model
$1.06
All tasks on the hard-case model
$50.00

List prices, before caching discounts and retries.

Get the cost plan

Add your work email and the plan opens on this page, built from your numbers. Print it or save it as a PDF for the budget owner.

  • Every model in the price list, priced for this workflow
  • 3 routing options, compared
  • A starting budget and daily cap for the gateway key
  • The report lines finance will ask for

How the cost is worked out

It follows the method in our guide to AI cost management. The prices come from that guide and every other figure from your inputs.

01

Provider usage

Tokens in and out per task, priced per million tokens for each model, split between the routine model and the hard-case model by the share you set.

02

Review and shared operation

Review minutes per task at the cost of a review hour, plus this workflow's share of the gateway, hosting and monitoring.

03

Cost per accepted result

The total divided by the results that pass review. Results that fail review still cost money, so they stay in the total.

Questions about prices, review time and caching

Where do the model prices come from?

Public list prices per million tokens, read from the providers' pricing pages on 7 September 2026 and listed in our guide to AI cost management. They change often and your contract may differ, so use them to see the spread between models.

Why count review time in an AI cost?

It's part of what an accepted result costs. In the starting example review time is the largest line, so a cheaper model barely moves the total.

Does it include prompt caching or retries?

No. Cached input is priced lower by several providers, which lowers the provider line. Retries and revisions add to it. If you know how often tasks are repeated, add the extra attempts to the tasks a month.

Set an AI budget for each team and see where it goes

One gateway for every AI call, with a budget per team and the cheapest model that passes each task. In one of our model trials, a cheaper model passed the task for $0.94 where the default model hit its $14 cap.

Ask your AI assistant about us

Open our summary in the assistant you use, or copy the prompt.

Read the prompt