What you can see
The dashboard surfaces three rolled-up time windows — last 24 hours, last 7 days, last 30 days — each showing:- Total requests across all your active keys.
- Error count and rate (any 4xx or 5xx response).
- Cache hit rate — what fraction of your traffic was served from the edge.
- Average latency measured at the Worker.
Latency caveat
Usage data is sampled and aggregated by Cloudflare Analytics Engine. Expect roughly one minute of lag between a request hitting the API and that request showing up on the dashboard. Don’t use this surface for real-time monitoring of incidents in progress — for that, instrument your own client.When to check
- During first integration. Confirm your requests are landing on the right endpoints with the right keys, and watch the cache hit rate climb as your client warms up.
- Before a launch. A dry-run sweep against staging, with the dashboard open, surfaces broken auth or off-by-one limit issues before users see them.
- After a deploy. Compare 1-hour totals before and after to confirm the change you shipped didn’t double your request volume.
- When approaching a quota. The portal flags any key over 80% of its monthly quota. If you see this, client-side caching is your first move.
Per-key vs aggregate
The dashboard rolls up across every active key on your account. If you mint separate keys per environment (recommended), you can read traffic patterns per key by filtering — e.g. development volume should look noisy and small, production volume should look smooth and large. Sudden inversions are usually a sign that production traffic is hitting your dev key by mistake.Programmatic access
We don’t expose usage data via the public API today. If you need programmatic access — for an internal dashboard, a billing reconciliation job, an alerting integration — write to support and tell us your use case. We’re tracking demand for a/v1/me/usage endpoint and your input shapes the priority.