api-bench

Benchmark your API endpoints right from the terminal. Sends N requests with C concurrency and gives you the numbers that matter: min/max/avg latency, p95, p99, RPS, and error rates.

Install

npm install -g @lxgicstudios/api-bench

Or run directly:

npx @lxgicstudios/api-bench https://api.example.com/health

Features

• Concurrency control - Set exactly how many parallel requests you want
• Latency percentiles - Get p50, p95, and p99 numbers
• RPS calculation - Requests per second throughput
• Error tracking - Status codes and error rate breakdown
• Compare mode - Benchmark two URLs side by side
• Ramp-up mode - Gradually increase concurrency to find breaking points
• POST support - Send custom bodies and headers
• JSON output - Pipe results to other tools
• Zero dependencies - Built with Node.js builtins only
• Progress bar - See what's happening in real-time

Usage

# Basic benchmark (100 requests, 10 concurrent)
api-bench https://api.example.com/health

# Heavy load test
api-bench https://api.example.com -n 1000 -c 100

# POST with JSON body
api-bench https://api.example.com/data -m POST \
  -b '{"name":"test"}' \
  -H 'Content-Type: application/json'

# Compare two endpoints
api-bench https://api-v1.example.com/users \
  --compare https://api-v2.example.com/users

# Ramp up concurrency gradually
api-bench https://api.example.com -n 500 -c 100 --ramp

# JSON output for CI/CD
api-bench https://api.example.com --json

Options

| Option | Description | Default | |--------|-------------|---------| | -n, --requests <num> | Total number of requests | 100 | | -c, --concurrency <num> | Concurrent requests | 10 | | -m, --method <method> | HTTP method | GET | | -H, --header <header> | Add header (repeatable) | - | | -b, --body <data> | Request body | - | | -t, --timeout <ms> | Request timeout | 10000 | | --compare <url> | Compare against second URL | - | | --ramp | Gradually increase concurrency | false | | --ramp-steps <num> | Number of ramp steps | 5 | | --json | Output results as JSON | false | | --help | Show help | - |

Output Example

 BENCHMARK RESULTS

  URL:          https://api.example.com/health
  Method:       GET
  Requests:     100
  Concurrency:  10
  Total Time:   2.34s

  Latency
  ────────────────────────────────────────
  Min:          12.3ms
  Max:          89.7ms
  Avg:          23.4ms
  Median:       21.1ms
  p95:          45.2ms
  p99:          78.3ms

  Throughput
  ────────────────────────────────────────
  RPS:          42.7 req/s
  Data:         125.4KB

Built by LXGIC Studios

GitHub | Twitter