anoman
Observability

Usage & Costs

See weighted tokens used against your monthly allowance, total spend, savings, and breakdowns by model, region and agent — in the dashboard or from the usage API.

Weighted tokens

Used vs your monthly allowance

The Usage & Costs page shows your weighted tokens used vs your monthly allowance with the reset date. Weighted tokens are the billing unit — see Quota & Overage for the formula and how caps work.

Because the multiplier scales with model class, cheaper and smaller models consume far fewer weighted tokens than flagship models — the usage breakdowns make that visible per model.

Usage API

Pull the numbers programmatically

Every dashboard figure has an API endpoint under /anoman/v1/usage:

EndpointReturns
/usageAccount totals — tokens, cost, and request count over a window.
/usage/by-modelSpend and tokens broken down per model.
/usage/by-regionSpend and tokens broken down per processing region.
/usage/by-agentSpend and tokens broken down per agent id.
/usage/savingsSavings from semantic cache hits and batch routing.
/usage/batch-economicsBatch rate, batch savings, and SLA compliance.
/usage/inout-splitInput vs output weighted tokens.
/usage/tier-infoYour tier plus its limits and allowance.
/usage/rate-limitsPer-window usage against your rate limits.
# Account totals — tokens, cost, requests over a window
curl "https://api.anoman.io/anoman/v1/usage" \
  -H "Authorization: Bearer anm-sk-..."

All endpoints use API-key auth and are scoped to your account.

How cost is metered

Metered from the provider, not estimated

Cost is metered from the provider's reported token usage on every call — never a local estimate. Weighted tokens are then derived from those raw tokens and the model's tier multiplier, and deducted from your allowance.

Related: For allowances, per-class caps, overage and burst credits, see Quota & Overage.

Watch your spend

Open the Usage page for live weighted-token and cost breakdowns.