API Monitoring Overview

Push-based monitoring for your own API endpoints.

What API monitoring is

API monitoring flips the model: instead of us probing your API, your application pushes metrics to us — latency, status codes, error flags and arbitrary custom metrics per request or in batches.

This answers questions an external check can't: real end-to-end latency inside your stack, per-endpoint error rates, payload-level metrics like queue depth or job duration.

How it works

  1. Create an API monitor in the dashboard (or POST /api/v1/uptime/api-monitoring/monitors flow via monitor type api)
  2. The monitor gets a dedicated push key — a bearer credential that only writes metrics to that one monitor
  3. Your app sends POST /api/v1/uptime/api-monitoring/push with the metrics
  4. If pushes stop within the expected interval, the monitor goes down — so a dead app is caught even when it stops reporting entirely

What you get

  • Latency percentiles and error-rate charts per endpoint
  • Custom metric widgets on your status page and dashboard
  • Uptime history like any other monitor — incidents, alerting, SLA all apply
  • Staleness detection: missing pushes mark the monitor unknown/down instead of silently looking healthy

Managing monitors

  • GET /api-monitoring/monitors/:id — config + state
  • PUT /api-monitoring/monitors/:id/config — update widgets, thresholds, expected interval
  • GET /api-monitoring/monitors/:id/metrics — recent pushed metrics
  • POST /api-monitoring/monitors/:id/regenerate-key — rotate the push key
NextPushing Metrics