Measure and compare API endpoint performance

REST API Benchmarking

Get precise latency percentiles (p50, p95, p99) for every REST endpoint. Compare performance across deployments, identify slow queries, and establish SLA baselines.
Standalone k6 example
Adapt this script and run it with k6 against an approved target. loadtest.dev accepts declarative GET/HEAD plans; it does not execute these scripts.
Percentile-based latency metrics (p50, p95, p99)
Endpoint-by-endpoint comparison
Throughput measurement (requests/second)
Error rate tracking
Custom thresholds and SLA validation
Historical comparison across test runs
benchmark.js
import http from 'k6/http';
import { check } from 'k6';
import { Trend } from 'k6/metrics';

const getUsers = new Trend('get_users_duration');
const getOrders = new Trend('get_orders_duration');

export const options = {
  vus: 100,
  duration: '5m',
  thresholds: {
    get_users_duration: ['p(95)<300'],
    get_orders_duration: ['p(95)<500'],
  },
};

export default function () {
  const users = http.get('https://api.example.com/users');
  getUsers.add(users.timings.duration);

  const orders = http.get('https://api.example.com/orders');
  getOrders.add(orders.timings.duration);
}

Frequently asked questions

Related guides

API Load Testing
Learn how to test HTTP APIs with standalone k6 scripts, measure latency and failed requests, and choose rates suited to your approved target.
HTTP Load Testing
Adapt a standalone k6 HTTP script with request methods, headers and JSON bodies, then inspect latency and failed requests on a target you own.
Performance Testing Checklist
Choose a workload, set latency and error thresholds, review resource metrics, and use standalone k6 checks to make performance results useful in CI.

Use an HTTP plan with loadtest.dev

For estate apps, declare GET/HEAD requests to a verified hostname and submit a plan through the authenticated API.
loadtest.dev — Part of the Narduk Enterprises network