Open-source performance software
1333 open-source libraries, tools and applications tagged performance, ranked by how closely performance is their primary purpose, then by adoption and maintenance health. Each entry links to a page with sources, use cases and when to choose or avoid it.
1333 products, primary matches first, then adoption-weighted; health v2 shown.
| Product | Health v2 | Stars | Maturity |
|---|---|---|---|
| chronoxor/CppServer CppServer is an ultra-fast, low-latency asynchronous socket server and client C++ library supporting TCP, SSL, UDP, HTTP, HTTPS, and WebSoc… | 81 | 1652 | stable |
| vk-com/kphp-kdb VK's KittenPHP/DB/Engine suite: a PHP-to-C++ translating compiler (KPHP) plus a set of specialized high-performance non-relational database… | 32 | 6065 | maintenance |
| intelxed/xed Intel XED is a C library for encoding and decoding X86 (IA32 and Intel64) machine instructions, converting between byte sequences and struc… | 89 | 1622 | stable |
| TraceMachina/nativelink NativeLink is a high-performance remote build cache and remote execution server written in Rust, speaking the open Remote Execution API. It… | 100 | 1621 | active |
| wuba/WBBlades WBBlades is a set of Mach-O based tools for iOS development that detects useless ObjC/Swift classes, protocols, and resources, analyzes pac… | 23 | 1608 | active |
| Frogging-Family/linux-tkg A set of shell scripts that automatically download, patch, and compile custom Linux kernels (version 5.4 and newer) with a selection of des… | 95 | 1605 | active |
| apache/gluten Apache Gluten is a middle-layer plugin that offloads JVM-based SQL engine execution (primarily Spark SQL) to high-performance native engine… | 88 | 1603 | active |
| pytorch/FBGEMM FBGEMM is a collection of highly optimized low-precision matrix multiplication and convolution kernels for server-side deep learning infere… | 91 | 1595 | active |
| rlerdorf/opcache-status A single-file PHP status page for Zend OPcache that provides an overview of cache configuration, cached scripts, and health metrics. It inc… | 71 | 1583 | active |
| tensorflow/model-optimization The TensorFlow Model Optimization Toolkit (tfmot) is a Python library providing tools to optimize machine learning models for deployment, i… | 79 | 1579 | stable |
| statoscope/statoscope Statoscope is a toolkit for analyzing and validating webpack bundle stats, providing a UI-based HTML report with full dependency trees, siz… | 57 | 1579 | active |
| rugarciap/Turbo-Boost-Switcher A macOS menu bar application that enables or disables Intel Turbo Boost by writing to the CPU's MSR register via a precompiled kernel exten… | 48 | 1558 | active |
| kffl/speedbump Speedbump is a TCP proxy written in Go that simulates variable, predictable network latency using configurable sine and sawtooth wave summa… | 23 | 1557 | active |
| JuliaLang/PackageCompiler.jl A Julia package for creating custom system images, standalone 'app' executables, and relocatable C libraries from Julia code. It moves Juli… | 95 | 1556 | active |
| benjamimgois/goverlay Goverlay is a Qt6 graphical application for configuring Linux gaming performance and overlay tools such as MangoHud, vkBasalt, vkSumi, and … | 100 | 1555 | active |
| nuxt/image Nuxt Image is an official Nuxt module providing plug-and-play image optimization for Nuxt applications. It offers drop-in <nuxt-img> and <n… | 85 | 1547 | stable |
| UCCL UCCL is a high-performance GPU communication library written in C++ that provides collectives (as a drop-in NCCL/RCCL replacement), P2P tra… | 69 | 1545 | active |
| VKCOM/kphp KPHP is a PHP compiler developed at VK.com that compiles a limited subset of PHP to C++ and then to a native binary running faster than pla… | 76 | 1526 | active |
| SimonCropp/Delta A .NET library that implements HTTP 304 Not Modified responses by leveraging database change tracking to generate ETags from last-updated t… | 77 | 1523 | active |
| ngtcp2/ngtcp2 ngtcp2 is a C library implementing the IETF QUIC transport protocol (RFC9000) with no external library dependencies for the core. It includ… | 94 | 1520 | active |
| ovg-project/kvcached kvcached is a Python library that brings OS-style virtual memory abstraction to KV cache management for LLM serving and training on shared … | 72 | 1509 | active |
| zouhir/jarvis Jarvis is a Webpack plugin that serves a browser-based dashboard showing bundle sizes, chunks, treeshakable vs CJS module imports, and asse… | 32 | 5478 | maintenance |
| Flipkart/recyclerlistview A high-performance list view component for React Native and web that uses cell recycling to reuse off-screen views, inspired by Android's R… | 35 | 5431 | maintenance |
| Shopify/semian Semian is a Ruby resiliency library that prevents cascading failures when accessing slow or unresponsive external services like databases a… | 84 | 1489 | stable |
| litespeedtech/openlitespeed OpenLiteSpeed is a high-performance, lightweight, open source HTTP web server developed by LiteSpeed Technologies under GPLv3. It uses an e… | 90 | 1483 | active |
| Lightning-AI/lightning-thunder Lightning Thunder is a source-to-source deep learning compiler for PyTorch that optimizes models for training and inference. It provides a … | 70 | 1475 | active |
| intel/compute-runtime Intel's open-source Graphics Compute Runtime (NEO) providing OpenCL 3.0 and oneAPI Level Zero compute API support for Intel HD Graphics and… | 98 | 1457 | active |
| timescale/tsbs TSBS is a Go-based benchmark suite for generating time series datasets and measuring read/write performance across many time series databas… | 66 | 1454 | active |
| tailcallhq/tailcall Tailcall is a high-performance GraphQL runtime written in Rust that lets you build GraphQL backends declaratively by annotating a GraphQL s… | 77 | 1443 | active |
| wix/import-cost Import Cost is a VSCode extension (plus a reusable Node module) that displays the size of imported packages inline in the code editor, usin… | 32 | 1382 | stable |
| hydro-project/hydro Hydro is a Rust framework for building correct and performant distributed systems using location-oriented programming, where a single progr… | 95 | 1379 | active |
| BytedTsinghua-SIA/CUDA-Agent CUDA-Agent is a large-scale agentic reinforcement learning system from ByteDance Seed and Tsinghua that trains LLMs to generate high-perfor… | 53 | 1338 | active |
| unum-cloud/UCall UCall is a high-performance JSON-RPC and REST serving library written in C with Python bindings, using io_uring and POSIX sockets to achiev… | 38 | 1335 | active |
| swow/swow Swow is a coroutine-based concurrent I/O engine for PHP, implemented as a C extension with most functionality exposed through PHP code. It … | 66 | 1330 | active |
| facebookincubator/AITemplate AITemplate is a Python framework that compiles deep neural networks into high-performance CUDA (NVIDIA) or HIP (AMD) C++ code for fast fp16… | 67 | 4730 | maintenance |
| tonybaloney/wily Wily is a command-line application that tracks and reports code complexity and other metrics across the git history of a Python project. It… | 83 | 1329 | active |
| MagicStack/httptools httptools is a Python binding for Node.js's fast HTTP parsers (llhttp and http-parser), providing request/response parser classes and a URL… | 90 | 1327 | stable |
| swaldman/c3p0 c3p0 is a mature, highly concurrent JDBC connection pooling library for Java, with support for caching and reuse of PreparedStatements. It … | 75 | 1317 | stable |
| theMackabu/ant Ant is a lightweight, high-performance JavaScript runtime built from scratch in C, featuring its own hand-built engine (Ant Silver) with a … | 84 | 1311 | active |
| savsgio/atreugo Atreugo is a high-performance, extensible micro web framework for Go built on top of fasthttp, with zero memory allocations in hot paths. I… | 23 | 1302 | active |
| PrunaAI/pruna Pruna is an open-source Python model optimization framework that makes AI models faster, smaller, cheaper, and greener via caching, quantiz… | 80 | 1288 | active |
| siddharthkp/bundlesize bundlesize is an npm CLI tool that checks whether your JavaScript bundle files stay under configured size limits, with gzip or brotli compr… | 62 | 4469 | maintenance |
| igorkasyanchuk/rails_performance A self-hosted Rails gem that monitors application performance, tracking response times (p50/p90/p99), throughput, slow requests, background… | 72 | 1247 | active |
| morganstanley/hobbes Hobbes is a strongly typed domain-specific language with an embedded JIT compiler and runtime for efficient dynamic expression evaluation, … | 76 | 1230 | active |
| microsoft/MInference MInference is a Microsoft library that accelerates long-context LLM inference using dynamic sparse attention, reducing pre-fill latency by … | 60 | 1229 | active |
| amireh/happypack HappyPack is a webpack plugin and loader that speeds up initial webpack builds by running file transformations (like babel-loader) in paral… | 10 | 4223 | maintenance |
| SimAI SimAI is a large-scale network simulation toolkit from Alibaba Cloud for modeling AI training and inference workloads on GPU clusters, publ… | 65 | 1201 | active |
| java-up-up/damai A high-concurrency online ticket booking system (modeled after Damai/大麦网) built with SpringCloud, SpringCloud Alibaba, Kafka, Redis, Sentin… | 67 | 1187 | active |
| rari-build/rari rari is a React Server Components framework whose HTTP server, RSC renderer, and router run on a Rust runtime with embedded V8, instead of … | 82 | 1176 | active |
| dropbox/goebpf A Go library for loading, attaching, and interacting with eBPF programs, maps, and perf events on Linux. It supports XDP, socket filters, k… | 69 | 1172 | active |
| Tencent/hpc-ops HPC-Ops is a production-grade C++/CUDA operator library for high-performance LLM inference, developed by Tencent's Hunyuan AI Infra team. I… | 58 | 1170 | active |
| web-infra-dev/rsdoctor Rsdoctor is a build analyzer for Rspack (and legacy webpack) projects that visualizes compilation time, loader/plugin behavior, bundle comp… | 90 | 1143 | active |
| davidfowl/BedrockFramework Bedrock Framework is a high-performance, low-level networking library for .NET that provides transport-agnostic APIs for building custom se… | 48 | 1134 | active |
| exAspArk/batch-loader A Ruby gem providing a generic lazy batching mechanism to avoid N+1 database or HTTP queries, inspired by Facebook's DataLoader and Haskell… | 54 | 1124 | stable |
| dimdenGD/ultimate-express A fast re-implementation of Express.js 4 built on µWebSockets, designed as a drop-in replacement with the same API. It validates compatibil… | 58 | 1108 | active |
| zalando-incubator/graphql-jit A TypeScript library that compiles GraphQL queries into optimized JavaScript functions using V8's JIT compilation, serving as a drop-in alt… | 95 | 1085 | active |
| Stonesjtu/pytorch_memlab A Python library providing line-level CUDA memory profiling and tensor inspection tools for PyTorch. It helps debug out-of-memory errors by… | 88 | 1080 | active |
| ngs-doo/dsl-json DSL-JSON is a high-performance JSON serialization library for the JVM (Java, Kotlin, Scala, Android) with compile-time databinding via anno… | 65 | 1073 | active |
| google/agi Android GPU Inspector (AGI) is a graphics profiling tool from Google for Android applications using Vulkan and OpenGL ES. It supports Qualc… | 61 | 1072 | active |
| cuOpt NVIDIA cuOpt is a GPU-accelerated optimization engine for linear programming, quadratic programming, mixed integer programming, and vehicle… | 83 | 1067 | active |
| epicweb-dev/cachified @epic-web/cachified is a TypeScript library that wraps any key-value store to act as a cache with TTL, stale-while-revalidate, parallel fet… | 89 | 1060 | active |
| tonbistudio/turboquant-pytorch A from-scratch PyTorch implementation of Google's TurboQuant (ICLR 2026) algorithm for compressing LLM key-value caches, including an impro… | 47 | 1047 | active |
| laruence/yaconf Yaconf is a PHP extension providing a persistent configuration container that parses INI files once at startup and serves them from immutab… | 95 | 1046 | active |
| PixelsCommander/HTML-GL HTML GL is a JavaScript library that renders HTML/CSS elements as WebGL textures to bypass slow DOM rendering, achieving high FPS animation… | 23 | 3520 | maintenance |
| Accelerate Accelerate is an embedded domain-specific language in Haskell for high-performance, multi-dimensional array computations. Programs written … | 59 | 1020 | active |
| eclipse-omr/omr Eclipse OMR is a set of open source C and C++ components for building robust, high-performance language runtimes across many hardware and O… | 95 | 1013 | active |
| brunofacca/active-record-query-trace A Rails plugin that logs a backtrace for every SQL query executed by Active Record in the development console and log. It helps developers … | 57 | 1009 | active |
| activej/activej ActiveJ is a self-sufficient Java platform built from the ground up, offering a modular suite of zero-dependency libraries for asynchronous… | 56 | 997 | active |
| zaiyunduan123/springboot-seckill A high-concurrency flash-sale (seckill) system built with SpringBoot, MySQL, Redis, RabbitMQ, and Guava. It demonstrates caching, rate limi… | 32 | 3291 | maintenance |
| pquerna/ffjson ffjson is a Go code generator that emits static MarshalJSON and UnmarshalJSON functions for your struct types, removing the runtime reflect… | 32 | 2991 | maintenance |
| facebook/wdt Warp speed Data Transfer (WDT) is an embeddable C++ library and command line tool for transferring files between two systems as fast as pos… | 67 | 2961 | maintenance |
| samccone/bundle-buddy Bundle Buddy is a tool that helps developers understand their JavaScript bundle size and why specific files end up in their bundles. It ana… | 32 | 2879 | maintenance |
| joe-bell/plaiceholder Plaiceholder is a server-side TypeScript library for generating low-quality image placeholders (LQIP) as colors, CSS gradients, SVG, or Bas… | 10 | 2316 | maintenance |
| Eclipse iceoryx Eclipse iceoryx is a C++ inter-process communication (IPC) middleware that provides true zero-copy data transfer via shared memory, with pu… | 87 | 2194 | maintenance |
| ixy network driver Ixy is an educational user-space network driver for Intel ixgbe 10 Gbit/s NICs, implemented in many programming languages (C, Rust, C++, Go… | 32 | 2185 | maintenance |
| qiujiayu/AutoLoadCache AutoLoadCache is a Java caching management library that decouples caching from business logic using AOP and annotations. It supports Redis,… | 23 | 2078 | maintenance |
| jiangwenyuan/nuster Nuster is a high-performance HTTP proxy cache server and RESTful NoSQL key-value cache server built on top of HAProxy, written in C. It is … | 23 | 1902 | maintenance |
| awslabs/llrt LLRT (Low Latency Runtime) is a lightweight JavaScript runtime built in Rust on the QuickJS engine, designed for fast-starting serverless a… | 88 | 8806 | experimental |
| basecamp/marginalia A Ruby gem that appends contextual comments (application, controller, action) to ActiveRecord SQL queries. It helps trace slow queries back… | 73 | 1769 | maintenance |
| squeaky-pl/japronto Japronto is a screaming-fast asynchronous HTTP toolkit with an integrated pipelining HTTP server for Python 3.5+, implemented in hand-tuned… | 32 | 8528 | experimental |
| sourcegraph/appdash Appdash is an application tracing system for Go, based on Google's Dapper and Twitter's Zipkin. It lets you instrument applications to reco… | 10 | 1705 | maintenance |
| idisfkj/android-startup An Android library for initializing application components at startup with explicit dependency ordering, sync/async execution, and multi-pr… | 23 | 1674 | maintenance |
| GoogleChromeLabs/size-plugin A Webpack plugin from GoogleChromeLabs that prints the gzipped sizes of your webpack assets and the changes since the last build. It tracks… | 58 | 1662 | maintenance |
| sindresorhus/gulp-rev A Gulp plugin that renames static assets by appending a content hash to their filenames (e.g., unicorn.css → unicorn-d41d8cd98f.css) for ca… | 68 | 1535 | maintenance |
| Swall0w/torchstat torchstat is a lightweight neural network analyzer for PyTorch that reports parameter counts, FLOPs, multiply-adds, and memory usage per la… | 23 | 1499 | maintenance |
| whackashoe/antimoderate AntiModerate is a tiny JavaScript library (under 2KB gzipped) for progressive image loading. It renders blurred micro thumbnails inline whi… | 32 | 1493 | maintenance |
| elementor/wp2static WP2Static is a WordPress plugin that generates a static copy of a WordPress site and deploys it to static hosting. It improves security, pe… | 23 | 1470 | maintenance |
| asfktz/autodll-webpack-plugin A webpack plugin that wraps DllPlugin and DllReferencePlugin to pre-bundle vendor dependencies without boilerplate, dramatically reducing b… | 23 | 1446 | maintenance |
| PeeHaa/OpCacheGUI A self-hosted web-based GUI for monitoring and managing PHP's OpCache. It provides a polished interface to view cache status, statistics, a… | 23 | 1444 | maintenance |
| Hermit Hermit is a Rust CLI tool that runs unmodified x86-64 Linux programs inside a hermetic sandbox, controlling sources of nondeterminism such … | 77 | 1417 | maintenance |
| robertknight/webpack-bundle-size-analyzer A utility that analyzes Webpack's JSON build output to show a tree of packages contributing to bundle size, ordered by module size. It can … | 32 | 1390 | maintenance |
| emmericp/ixy ixy is a userspace packet processing framework and network driver for Intel 10 Gbit/s NICs (82599/ixgbe family) and virtio, written from sc… | 32 | 1319 | maintenance |
| 3rd-Eden/memcached A fully featured Memcached client library for Node.js built with scaling and high availability in mind. It supports Memcached clusters via … | 23 | 1313 | maintenance |
| tysam-code/hlb-CIFAR10 A single-file PyTorch implementation that trains a neural network to 94% accuracy on CIFAR-10 in under 6.3 seconds on a single A100 GPU, fo… | 23 | 1310 | maintenance |
| chengzeyi/stable-fast Stable Fast is an ultra lightweight performance optimization framework for HuggingFace Diffusers pipelines on NVIDIA GPUs. It achieves SOTA… | 21 | 1305 | maintenance |
| c9s/Pux Pux is a high-performance PHP routing library with an optional C extension, offering Sinatra-style PCRE patterns, controller auto-mounting,… | 32 | 1263 | maintenance |
| session-replay-tools/cetus Cetus is a high-performance MySQL proxy middleware written in C that provides transparent routing between applications and backend MySQL se… | 23 | 1241 | maintenance |
| ANS (Accelerated Network Stack) ANS is a DPDK-native userspace TCP/IP stack written in C, based on the FreeBSD implementation, providing zero-copy mbuf handling and a BSD-… | 23 | 1218 | maintenance |
| Netflix/rend Rend is a memcached-compatible proxy server written in Go that manages data chunking and coordinates L1 (RAM) and L2 (SSD-backed) cache tie… | 32 | 1192 | maintenance |
| Rust-GPU/rust-cuda An ecosystem of Rust libraries and tools for writing and executing fast GPU code fully in Rust, targeting NVIDIA CUDA. It lets developers a… | 66 | 5399 | experimental |