Sharkbench

Benchmarking programming languages and web frameworks

Quarkus: Imperative vs Reactive

Quarkus is a Kubernetes-native Java framework tailored for GraalVM and OpenJDK, crafted from best-of-breed Java libraries and standards. It offers a unified reactive and imperative programming model optimized for containers and serverless environments. Quarkus focuses on fast startup time, low memory footprint, and developer productivity.

This benchmark tests how fast a framework can perform concurrent HTTP requests, I/O operations, and JSON de/serialization.

OS: Linux/DockerCPU: Ryzen 7 7800X3DLast Updated: 2025-05-04

Comparison

Reactive processes 1.11x requests per second compared to Imperative, a 11.39% improvement.

Mode	Requests per second Indicates how many requests per second the framework can handle. (higher is better)	Latency Indicates how long it takes to process a request. (lower is better)	Stability Indicates how stable the framework is under load: median / P99 latency (higher is better)	Memory Indicates how much memory (RAM) the framework uses. (lower is better)
Reactive	6240 (+11.39 %)	0.7 ms (-5.31 %)	0.82 % (-0.04 %)	362.5 MB (-7.12 %)
Imperative	5602	0.8 ms	0.86 %	390.3 MB

All results

MedianP99

Framework	Requests per second Indicates how many requests per second the framework can handle. (higher is better)	Latency Indicates how long it takes to process a request. (lower is better)	Stability Indicates how stable the framework is under load: median / P99 latency (higher is better)	Memory Indicates how much memory (RAM) the framework uses. (lower is better)
Java (OpenJDK v21) / Quarkus v3 (Reactive)	6240	0.7 ms	0.82 %	362.5 MB
Java (OpenJDK v21) / Quarkus v3 (Imperative)	5602	0.8 ms	0.86 %	390.3 MB
Java (OpenJDK v17) / Quarkus v3 (Reactive)	6096	0.7 ms	0.83 %	341.5 MB
Java (OpenJDK v17) / Quarkus v3 (Imperative)	5694	0.8 ms	0.85 %	370.6 MB

Popular web frameworks

How does Quarkus compare to other popular frameworks?

MedianP99

Framework	Requests per second Indicates how many requests per second the framework can handle. (higher is better)	Latency Indicates how long it takes to process a request. (lower is better)	Stability Indicates how stable the framework is under load: median / P99 latency (higher is better)	Memory Indicates how much memory (RAM) the framework uses. (lower is better)
Rust / Axum	21030	1.6 ms	71.96 %	8.5 MB
C# (.NET) / ASP.NET Core	14707	1.2 ms	2.64 %	136.5 MB
Java (OpenJDK) / Quarkus (Reactive)	6240	0.7 ms	0.82 %	362.5 MB
JavaScript (Node.js) / Express	5766	5.5 ms	64.50 %	82.5 MB
Java (OpenJDK) / Quarkus (Imperative)	5602	0.8 ms	0.86 %	390.3 MB
Go / Gin	3546	1.0 ms	1.14 %	16.7 MB
Python / FastAPI (Uvicorn)	1185	21.0 ms	21.21 %	41.2 MB
Java (Temurin) / Spring Boot (MVC)	1105	1.7 ms	1.67 %	596.8 MB
Python / Flask (Gunicorn)	1092	7.7 ms	9.22 %	90.3 MB
Python / Django (Gunicorn)	950	8.8 ms	10.33 %	130.1 MB
PHP / Laravel	299	101.7 ms	56.47 %	84.2 MB