Benchmark GitHub Repositories (2026) · Open Source Tools

Cista is a simple, high-performance, zero-copy C++ serialization & reflection library.

A deep learning toolkit for healthcare AI: EHR (MIMIC, eICU, OMOP), clinical text, signals and imaging. 30+ models, reproducible pipelines.

Fast Compiler for C# Expression Trees and the lightweight LightExpression alternative. Diagnostic and code generation tools for the expressions.

ParseBench - A Document Parsing Benchmark for AI Agents

A modern iperf3 alternative with a live TUI, multi-client server, and QUIC support. Built in Rust.

Advanced System Analysis Software

Swift benchmark runner with many performance metrics and great CI support

Research infra for creating RL environments, post-training, and evals.

A framework for reproducible, comparable benchmarks

[Public preview] Externally scored agentic ML research benchmark: 60 tasks, real competition ground truth. Open protocol, operated evaluation.

Rust libraries and programs focused on succinct data structures

The first exam for code agents in the physical world: frozen robot tasks from many benchmark suites, each answered with a program and verified in a separate sandbox. Runs on Harbor.

glTF 2.0 versions of common graphics research scenes

aws-bench measures how well AI agents and model combinations perform on real AWS work — diagnosing misconfigurations, provisioning infrastructure, and operating live cloud environments.

Playing around "Less Slow" coding practices in Rust, from numerical micro-kernels to coroutines, ranges, and polymorphic state machines

:sunrise: A comprehensive, cutting-edge open-source framework for next-generation Transformers, with compute-matched benchmarking and a platform for autonomous architecture research.

Set of benchmarks for the Ruby programming language

Benchmarks for bundlers and build tools, including Rspack, Rsbuild, webpack, Vite, Rolldown, esbuild, Parcel, Farm and Utoo.

🧪Yet Another ICU Benchmark: a holistic framework for the standardization of clinical prediction model experiments. Provide custom datasets, cohorts, prediction tasks, endpoints, preprocessing, and models. Paper: https://arxiv.org/abs/2306.05109

Benchmark local LLM serving by replaying real agent trajectories