Use case
How to bill customers for their AI usage
Usage-based billing for AI features needs a real, attributed cost per customer as its input, not an estimate.
Per-customer attributed cost, produced by tagging every proxied call with x-customer-id, is the input usage-based billing needs. Spendline records and attributes the cost; it does not itself issue invoices to your end customers, that stays in your own billing system.
How Spendline does this
The same ledger entry used for cost attribution and margin reporting is the source of truth for what a customer actually consumed. Rather than estimating usage from a proxy metric, the exact resolved cost of every call attributed to that customer is available to feed into whatever billing system charges them, usage-based, a markup on cost, or a hybrid plan.
Why it matters
Billing a customer for AI usage on an estimate, a flat per-seat guess, or a rough token count without the real resolved price, either undercharges on the customers actually driving cost or overcharges the ones who are not. An attributed, per-call cost is the difference between a billing model that holds up and one that quietly loses margin on your heaviest users.
Frequently asked questions
Does Spendline send invoices to my customers?
No. Spendline attributes and records the cost. Turning that into an invoice, whether usage-based or a flat markup, happens in your own billing system.
Can I mark up the attributed cost before billing it?
The attributed figure is your raw AI cost per customer; how you price that to your customer, at cost, with a margin, or bundled into a plan, is a business decision made downstream of the number Spendline provides.
See where this stands in your own setup
This page describes the mechanism. The 5 minute assessment scores your own attribution, enforcement, and reconciliation setup, so you know exactly which of these problems you actually have today.
Run the 5 minute assessment