Ideal API Response Time
Benchmarks and alert thresholds for your APIs.
An API's response time directly impacts user experience and application performance. Slow APIs create frustration and can cascade into outages.
This guide presents benchmarks by API type and recommended alert thresholds.
Benchmarks by API Type
Recommended response times by context:
| API Type | Excellent | Acceptable | Slow |
|---|---|---|---|
| Simple REST API | < 100ms | 100-300ms | > 500ms |
| API with DB | < 200ms | 200-500ms | > 1s |
| Aggregation API | < 500ms | 500ms-1s | > 2s |
| Export/Report API | < 2s | 2-5s | > 10s |
Influencing Factors
- Request complexity: simple requests should be fast, aggregations can be slower.
- Geographic distance: network latency adds to processing time.
- Server load: times increase under heavy load.
- Dependencies: each external call adds latency.
Configure MoniTao Alerts
- Warning threshold: alert when time exceeds the usual 95th percentile.
- Critical threshold: alert immediately beyond the "slow" threshold.
- Trends: monitor gradual degradation before thresholds.
Frequently Asked Questions
Should I alert on every slow request?
No, use a threshold of N slow requests in X minutes to avoid false positives.
Which metric to use: average or percentile?
P95 or P99 is more representative than average, which hides outliers.
How to improve response times?
Optimize DB queries, add caching, reduce synchronous dependencies.
Does MoniTao measure TTFB?
Yes, response time includes TTFB (Time To First Byte).
Ready to Sleep Soundly?
Start free, no credit card required.