Attribution & Finance
Credit burndown
Drawing down a customer's prepaid balance as they consume, and recognizing that revenue on burn rather than at purchase. Prepaid credits are deferred revenue when sold; each metered unit of usage converts a slice of that liability into recognized revenue — which is why accurate metering is also an accounting control, not just a billing one.
Example
A tenant prepays for 50M tokens. Each call burns down the balance and recognizes that portion of the prepayment as earned revenue. When the balance runs low, an alert fires before service is interrupted — and the deferred-revenue liability on the books matches the credits still unused.
This is a Tokenality concept. See how it works in the product overview or the live playground.
Related terms
Usage-based billing
Charging customers in proportion to what they actually consume — tokens, calls, or compute — instead of a flat fee, seat count, or user-count proxy. It aligns price with cost: because AI cost scales with tokens, usage-based billing is the only pricing that keeps a reseller's margin stable as customers grow.
ASC 606 (for AI usage)
The revenue-recognition standard that requires you to recognize revenue as a performance obligation is satisfied. For usage-based AI, that means recognizing revenue as tokens are consumed — not when a contract is signed or credits are sold. Prepaid credits are deferred revenue drawn down on burn; a spreadsheet can't defend that treatment to an auditor.
Billable tokens
The tokens you actually charge a customer for — as opposed to the tokens you merely consumed. When you resell AI, billable tokens are the input and output tokens attributed to a specific customer of record, rated at your price (your provider cost plus your margin) rather than the raw provider cost.