Monitor API Usage & Set Quota Alerts
Once you are calling the Public REST API in production, you need to see how much of your quota you are burning and be warned before you hit a wall. The API Keys area does both: a real-time Usage dashboard against your plan limits, per-token hard caps for fail-safe blocking, and email quota alerts at thresholds you choose.
Before you start#
Open Settings → API Keys. It has three tabs — Keys, Usage and Alerts — and the active tab is remembered in the URL so deep links work. Public API access is plan-gated:
- Free — not included. The page shows a lockout screen ("The Public REST API isn't included in your plan") instead of the tabs.
- Growth — 2,500 calls/day and 25,000/month.
- Advance — 5,000 calls/day and 50,000/month.
If you haven't generated a token yet, start with API Keys: Build Custom Integrations, then come back here to watch consumption.
Read the Usage dashboard#
The Usage tab tracks consumption in near real time. Use the period selector at the top — Last 24 hours, Last 7 days, Last 30 days, This month or Last month — to drive the chart and breakdowns.
- Progress bands — two cards, Monthly and Today, each showing calls used, the plan limit, the percent consumed, and when the quota resets. A projection chip estimates your month-end total and flips to a red "Will exceed plan" warning when you're trending over.
- Data refreshed… — a freshness line stating how recently the figures updated, with a "Configure notifications" link into the Alerts tab.
- Time-series chart — call volume over the selected period.
- Breakdown — one panel with four tabs: By Token, By Endpoint, By IP and By Status, so you can see which key, route, source or response code is driving usage.
- Recent activity (last 100 calls) — a live log with time, token, method, endpoint, status code and latency for debugging.
On a Custom plan with unlimited limits the progress bars are replaced by an "Unlimited (Custom plan)" banner; on a paid plan with no calls yet you'll see "No API calls this month yet".
Set hard caps per token#
By default, paid plans use soft-mode overage: calls keep succeeding past your quota, and the overage is logged and shown in the dashboard. When you'd rather a specific token stop dead at the limit, open its Hard-cap settings from the Keys tab:
- Block at daily limit — the token returns HTTP 429 once its daily quota is reached.
- Block at monthly limit — the token returns HTTP 429 once its monthly quota is reached.
Saving confirms with "Hard-cap settings updated." Hard caps are ideal for an untrusted third-party automation token, where limiting blast radius matters more than uninterrupted continuity.
Create quota alerts#
The Alerts tab sends email when usage crosses a threshold. Press Add custom alert and set:
- Alert type — Monthly quota or Daily quota.
- Threshold (%) — an integer from 1 to 200. Values above 100 alert on overage (for example, 110 means 10% over your quota).
- Recipients — one or more comma-separated email addresses; at least one is required.
- Recurring — Every period re-arms the alert each day/month, while Once never re-arms after it first fires.
System-default alerts have a fixed type, threshold and recurrence — you can change their recipients or disable them, but not their type or threshold.
Good to know#
- Usage figures lag slightly behind live traffic — the freshness line tells you exactly how current the numbers are, so treat the very latest calls as approximate.
- Threshold percentages accept overage values on purpose; use 100+ if you only want to hear about breaches, not approaches.
- Hard caps and alerts are complementary — alerts warn you, hard caps enforce a stop. Many teams run alerts everywhere and hard caps only on risky tokens.
Ready to implement this?
OpenAI Engage and apply what you just learned to your own workspace.
Open AI Engage CRM