A little utility for doing side-by-side benchmarks in nodejs

<h1 align="center"> <img src="https://raw.githubusercontent.com/RafaelGSS/bench-node/refs/heads/main/assets/logo.svg" alt="Bench Node logo" /> Bench Node </h1>

A React component to run and visualize `@probe.gl/bench` benchmarks.

Model-independent agentic benchmark harness for WrongStack (bundled smoke suite, Aider polyglot, SWE-bench Verified) with deterministic graders, harness fingerprinting, and run-to-run comparison reports.

Flow Bench

Run the webgpu-bench GPU benchmark suite from the command line, no browser needed.

Freeze the model, evolve the harness. Two measured applications: (1) SWE-bench code-repair — conformant GLM->Opus empty-patch cascade resolves 51.3% Lite (n=300) and 55.6% Verified (278/500, Wilson 95% CI [51.2, 59.9], official swebench gold eval, no gold

Web player for t-bench session recordings.

ESLint rules for Vite, Vitest, and Vitest bench configuration and runtime patterns.

WebGL performance monitor that showing percentage of GPU/CPU load

In-memory test bench (createBench) for @sanity/workflow-engine.

bench-rest - benchmark REST (HTTP/HTTPS) API's. Node.js client module for easy load testing / benchmarking REST API' using a simple structure/DSL can create REST flows with setup and teardown and returns (measured) metrics.

Bahulam Code — abundance, in your terminal. CLI-first, reliability-first, sub-agents, 65.6% SWE-bench Verified.

A local-only Ollama inference benchmark implementing the osai-bench measurement protocol.

Benchmark harness.

Run benchmark with Jest

Retrieval latency ladder benchmarks + CI regression gates for @remnic/core

Micro bench!

Deepkit Bench

Evolution harness: observe failures, mutate, re-bench, select — human-gated promotion, offline lineage bundles.