Skip to content

Performance & Benchmarks

Velociradix is engineered for minimum latency and maximum requests per second.


📊 Benchmark Results

Benchmarked on Apple M1 (ARM64) using autocannon with 500,000 requests, 100 concurrent connections, and 10 HTTP pipelining:

Engine / FrameworkGET /json (RPS)Avg LatencyRank
🚀 Velociradix (fastGet)114,490 req/s8.26 ms🥇 #1 Winner
Fastify41,605 req/s23.55 ms🥈 #2
Velociradix (Standard)19,915 req/s50.26 ms🥉 #3 (nearly 2x Express)
🐌 Express11,503 req/s88.04 ms#4

💡 Tip: Remember to disable console logging (logger()) during production benchmark runs to prevent stdout I/O bottlenecks and achieve maximum throughput.


⚡ C++ Fast-Path (fastGet, fastPost, fastRoute)

When registering routes via app.fastGet(), static or pre-calculated JSON responses are served directly from native C++ memory, completely bypassing the Node.js V8 JS event loop and avoiding GC overhead.

js
// Served at 120,000+ req/s directly in native C++ memory
app.fastGet('/health', { status: 'ok', engine: 'velociradix' });
app.fastGet('/ping', 'pong');

Released under the MIT License.