BillingUsage Metering

Fine-grained AI usage metering, tied to customers

Track AI and vector usage across providers, clouds, and services. Normalize it into a clean event ledger so product, finance, and ops can make decisions with real unit economics.

Attribution

Customer and SKU mapping

Every event ties back to the account, workspace, API key, endpoint, and the SKU that drove it.

Coverage

Across providers and infra

LLMs, vector stores, storage, and internal services recorded under one schema.

Durability

Ledger you can trust

Immutable, queryable records designed to hold up when finance and ops ask hard questions.

What we meter

The details you need to price fairly and debug spikes

Enough granularity to drive billing and product decisions without turning into a dashboard graveyard.

Model calls
requests, tokens, input and output sizes, model variants
Vector operations
upserts, queries, deletes, index usage
Storage
reads, writes, bytes stored, retention windows
Compute signals
internal service cost signals and shared infra usage
Attribution keys
customer, workspace, API key, endpoint, SKU, feature flags

How it works

Meter → Normalize → Attribute → Export

A clean pipeline that keeps engineering fast and keeps finance sane.

Step 1

Instrument

Capture model calls, tokens, vector queries, storage reads and writes, plus internal service signals.

Step 2

Normalize

Translate provider-specific units into a stable event schema and compute cost consistently.

Step 3

Attribute

Map events to customer, workspace, API key, endpoint, and SKU with strong ownership rules.

Step 4

Export

Send usage to billing, data warehouses, or invoices while keeping an immutable ledger.

Why it matters

Metering is the foundation for billing and control

Without metering, you guess. With metering, you can price, limit, audit, and reconcile.

Pricing

Usage-based plans that stay profitable

Price per unit, set margins, and stop subsidizing heavy customers without realizing it.

Product

Feature and model decisions with real cost

Know what each workflow costs before you roll it out to every customer.

Ops

Investigate spikes without hand-waving

Trace usage back to customers and endpoints when spend changes fast.

Finance

Records that hold up

Usage history that supports invoices, reconciliation, and month-end workflows.

Stop guessing who your AI spend belongs to

If you are shipping AI to customers, you need clean metering before pricing and billing can work. BirdyFoot gives you a usage ledger you can build on.