A feature matrix does not survive a finance review. The internal case for an LLM gateway is four measurable outcomes — total spend, unit cost, performance under load, and shipping speed — each with a number attached.
Request count and dollar cost tell different stories, and the gap between them is where the savings are. Here are the four shapes an overlay of cost and usage produces, which one to chase first, and what makes the numbers trustworthy enough to act on.
Your balance is not a stored number, it is the sum of a ledger. Here is how to read the four numbers on the balance card, the entry types behind them, and the two identities that prove your balance is doing what it should.
"The AI bill went up" becomes a query once spend carries structure. Three attribution layers — virtual keys, teams, and the OpenAI-spec user field — turn one opaque total into a breakdown you can group, filter, and cap.
If you resell AI, the model bill is the wrong unit. Attribute every call to the customer it served with the OpenAI user field, cap each customer independently, and reconcile the sum against your ledger to the cent.
Provider pricing does not normalize on its own — per-token, per-image, per-second, provisioned. Here is how nRouter turns that into one settled cost per request that your app, your ledger and your dashboard all read, and why an unknown cost is reported as absent rather than as zero.
