What's the difference between showback and chargeback?

Showback reports to a team or customer what their AI usage cost, without moving any money. Chargeback bills them for it, debiting a cost centre or adding it to an invoice. Showback usually comes first in practice: teams see their number for a few cycles and correct the obvious waste before anyone is charged against it.

How do I give each team visibility into its own AI spend?

Attribute every call to a team identifier at the point it is made, then report per-team totals on the team's own cadence rather than only in an aggregate. The prerequisite is that attribution coverage is high enough to trust — a large unattributed bucket destroys confidence in every team's number. Showback works when a team can see both its total and which of its agents drove it.

Chargeback and showback: definition

Why it matters

Both are standard FinOps practice for cloud spend, and LLM spend is arriving at the same treatment — with the complication that the cost-causing unit is an agent rather than a server, and one API key typically serves many cost centers at once. Showback usually comes first: teams see their number for a few cycles before anyone is charged against it.

Example

A platform team runs shared agent infrastructure for six product teams. Showback sends each team a monthly statement of the spend its agents caused; chargeback debits their cost centers for it. Either way the prerequisite is the same: spend attributed per team, reliably enough for finance to act on.

How it's measured

By the completeness and auditability of the attribution feeding it — unattributed spend has to land somewhere, and how large that bucket is determines whether finance trusts the report. Capsera carries team, customer, and cost-center identifiers on events for exactly this decomposition.

Questions this page answers

What's the difference between showback and chargeback?: Showback reports to a team or customer what their AI usage cost, without moving any money. Chargeback bills them for it, debiting a cost centre or adding it to an invoice. Showback usually comes first in practice: teams see their number for a few cycles and correct the obvious waste before anyone is charged against it.
How do I give each team visibility into its own AI spend?: Attribute every call to a team identifier at the point it is made, then report per-team totals on the team's own cadence rather than only in an aggregate. The prerequisite is that attribution coverage is high enough to trust — a large unattributed bucket destroys confidence in every team's number. Showback works when a team can see both its total and which of its agents drove it.

Related terms

Cost attribution — Cost attribution is the assignment of every unit of LLM spend to the agent, team, feature, or customer that caused it, rather than leaving it aggregated under a shared API key.
Per-agent budget — A per-agent budget is a spending envelope scoped to one named agent, rather than to the API key, project, or account the agent shares with everything else.
Cost per run — Cost per run is the total LLM spend attributed to one execution of an agent task, summed across every call, model, and agent that execution triggered.