Usage & quotas
Every action APIXX takes on your behalf is metered. The Usage page is the single place to see consumption against your plan, broken down by meter, API key, and time.
What we meter
| Meter | Unit | Counts |
|---|---|---|
| Transactions | successful run-record pairs | One per record successfully written to a destination. |
| API calls | HTTP request to apixx.io | Counted per endpoint; read and write tracked separately. |
| Ingestion volume | MB of source payload read | Net of compression. |
| Active connectors | distinct profiles with traffic in window | Used for tier-based billing. |
| Canonical records | rows stored in APIXX Data | Charged monthly average, not peak. |
Time windows
The Usage page supports 24h, 7d, and 30d windows. Switching the window re-scopes every chart on the page. Hour-of-day and day-of-week breakdowns help you spot consumption patterns — for example, weekend backfills that you didn't realize were running.
Per-API-key activity
Every API key is tagged on every request. The keys panel shows per-key call volume, last-used timestamp, and rolling error rate. Idle keys older than 90 days are flagged for review — rotate or revoke them.
Quota & overage
Plan limits are soft by default — Flows keeps running but raises a warning banner at 80% and an alert at 100%. Sustained overage is reconciled in your next invoice. Your contract defines whether hard caps apply.
Ingestion volume
We measure the raw bytes read from each source connector before any transform. Large product catalogs and historical backfills dominate this meter. Tips to reduce:
- Use updated_since filters on every read flow.
- Project only the fields you need with Flow → Settings → Source projection.
- Run one-off backfills on off-peak schedules.
Exporting usage
Download a CSV of the current window from the Usage page, or hit GET /v1/usage?from=&to=&granularity=hour via the API for programmatic access. Useful for cost-allocation to internal teams.