Aspectran Ranks #1 Among Java Web Frameworks in Global Benchmark

Aspectran has officially been registered and evaluated in the open-source Web Frameworks Benchmark project.

In the latest benchmark results released today, Aspectran achieved #1 in Average Throughput (Average RPS) and #1 across Concurrency 64 and 256 among all tested Java web frameworks.

Competing against 353 web frameworks across various programming languages (including native languages like C++, Rust, and Go), Aspectran secured a position in the top 12% globally (Rank 44 / 353), demonstrating industry-leading performance.

📊 Web Frameworks Benchmark Overview

The Web Frameworks Benchmark is a reputable open-source benchmarking suite that measures raw HTTP request handling throughput under standardized, isolated container environments.

• Tested Endpoints:
  • GET / (Plain root request)
  • GET /user/${id} (Route parameter extraction and plain text response)
  • POST /user (Basic payload handling)
• Tested Concurrency Levels: 64, 256, and 512 concurrent connections
• Runtime Environment: Java 21, Linux isolated container

🏆 Java Web Frameworks Benchmark Results

Below are the detailed performance metrics for all 21 Java web frameworks running on Java 21, sorted by Average RPS.

🌟 Key Performance Highlights

1. Superior Immediate Responsiveness from Initial Load

While many enterprise frameworks experienced cold starts or thread pool allocation bottlenecks at 64 concurrency (yielding 2,000–6,000 req/s), Aspectran delivered 34,305 req/s immediately at 64 concurrency, outperforming all other Java frameworks from the very first test tier.

As concurrency increased to 256 and 512, throughput scaled steadily to 36,482 req/s and 37,807 req/s, exhibiting rock-solid stability without request latency spikes.

2. Synergy Exceeding the Underlying Undertow Engine

Aspectran runs on embedded Undertow via aspectran-with-undertow. While high-level application frameworks typically introduce latency over the raw server engine, Aspectran posted an average throughput of 36,198.0 req/s, surpassing raw Undertow’s 35,741.3 req/s. This demonstrates that Aspectran’s LightRequestHandlerFactory and translet execution pipeline introduce zero noticeable runtime overhead and extract maximum capability from the underlying NIO engine.

3. More Than Double the Throughput of Mainstream Frameworks

Aspectran delivered more than 2.0x to 2.4x the throughput of Spring (17,394 req/s) and Spring WebFlux (15,039 req/s), and 1.3x to 1.5x higher throughput than cloud-native stacks such as Quarkus (26,596 req/s) and Micronaut (23,874 req/s).

💡 How Aspectran Delivers Both Performance and Enterprise Capability

In web framework design, it is often assumed that “comprehensive enterprise frameworks are inherently heavy, and fast frameworks are merely barebones routing libraries.”

Aspectran was purposefully architected from its inception to break this compromise: delivering deep enterprise-grade capabilities alongside raw, engine-level throughput.

1. Translet-Based Shortest Path Execution

At the core of Aspectran is the Translet, an execution unit that links request reception, parameter mapping, action execution, transformation, and response rendering in a direct, streamlined pipeline. By removing multi-layered filter chains and redundant object allocations, requests travel through the shortest execution path.

2. Zero-Overhead ActivityContext & Lightweight IoC/AOP

Rather than incurring continuous runtime reflection penalties and heavy dynamic proxy lookup costs, Aspectran operates on a pre-compiled, optimized ActivityContext. Enterprise features such as IoC/DI and non-invasive AOP are applied with minimal memory allocation during active request cycles.

3. A Full-Featured Enterprise Framework, Not a Stripped-Down Router

What makes this achievement particularly meaningful is that Aspectran was not stripped of features for this benchmark. Aspectran competed with its complete runtime stack—including its full IoC/DI container, declarative transaction management, task scheduling (Scheduler), background daemons (Daemon), session clustering, APON data serialization, and multi-format template engines.

🎯 Looking Forward

These benchmark results validate years of rigorous architectural refinement and optimization.

Aspectran will continue to evolve, empowering developers to build high-throughput cloud-native services and mission-critical enterprise systems with uncompromising performance and high developer productivity.

You can inspect Aspectran’s benchmark setup and configuration files directly in the Web Frameworks GitHub Repository.