Open-source gpu-computing software
544 open-source libraries, tools and applications tagged gpu-computing, ranked by how closely gpu-computing 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.
544 products, primary matches first, then adoption-weighted; health v2 shown.
| Product | Health v2 | Stars | Maturity |
|---|---|---|---|
| GeeeekExplorer/nano-vllm A lightweight vLLM-style LLM inference engine implemented from scratch in about 1,200 lines of Python. It offers fast offline inference wit… | 51 | 15834 | active |
| FreeToken FreeToken is an edge-native Mixture-of-Experts (MoE) LLM serving engine that runs frontier-scale open-weight models on consumer hardware by… | 79 | 14561 | active |
| facebookresearch/vggt VGGT (Visual Geometry Grounded Transformer) is a feed-forward transformer model from Meta AI and Oxford VGG that infers 3D geometry—camera … | 55 | 14499 | active |
| Eclipse Deeplearning4j Eclipse Deeplearning4J is an open-source deep learning framework and ecosystem for the JVM, including the ND4J linear algebra library, the … | 76 | 14269 | active |
| Open3D Open3D is an open-source C++ and Python library for 3D data processing, offering data structures, algorithms, and pipelines for point cloud… | 86 | 14097 | active |
| apache/tvm Apache TVM is an open machine learning compiler framework that takes pre-trained models and compiles them into optimized, deployable module… | 95 | 13840 | active |
| openwall/john John the Ripper jumbo is an open-source offline password cracker supporting hundreds of hash and cipher types, from Unix and Windows passwo… | 71 | 13668 | active |
| NVIDIA/TensorRT NVIDIA TensorRT is an SDK for high-performance deep learning inference on NVIDIA GPUs, comprising an inference compiler that converts train… | 93 | 13362 | stable |
| modelscope/DiffSynth-Studio DiffSynth-Studio is an open-source diffusion model engine from the ModelScope community that integrates mainstream image, video, and audio … | 77 | 13229 | active |
| cupy/cupy CuPy is a NumPy/SciPy-compatible array library for GPU-accelerated computing with Python, running on NVIDIA CUDA or AMD ROCm. It acts as a … | 92 | 12362 | stable |
| LMCache/LMCache LMCache is a KV cache management layer for LLM inference that stores, compresses, and reuses KV caches across requests, sessions, and servi… | 88 | 12133 | active |
| Triton Inference Server NVIDIA Triton Inference Server is an open-source inference serving software that deploys AI models from multiple frameworks (TensorRT, PyTo… | 98 | 11075 | stable |
| OpenVINO OpenVINO is an open-source toolkit from Intel for optimizing and deploying deep learning inference across CPU, GPU, and NPU hardware. It su… | 95 | 10996 | stable |
| cumulo-autumn/StreamDiffusion StreamDiffusion is a Python pipeline for real-time interactive diffusion-based image generation, achieving 100+ fps on modern GPUs. It opti… | 18 | 10817 | active |
| SkyPilot SkyPilot is an open-source AI compute platform that unifies fragmented infrastructure (Kubernetes, Slurm, VMs, 20+ clouds) into a single po… | 100 | 10724 | active |
| facebookresearch/xformers xFormers is a PyTorch-based library of hackable, optimized Transformer building blocks with custom CUDA kernels for fast, memory-efficient … | 87 | 10565 | active |
| deepseek-ai/DeepEP DeepEP is a high-performance GPU communication library for expert parallelism (EP) in MoE training and inference, providing high-throughput… | 55 | 10263 | active |
| OpenRLHF/OpenRLHF OpenRLHF is a high-performance, production-ready open-source RLHF framework built on Ray + vLLM + DeepSpeed for scalable reinforcement lear… | 89 | 10040 | active |
| BentoML BentoML is a Python framework for building online serving systems for AI apps and model inference, turning model inference scripts into RES… | 91 | 8909 | stable |
| DeepDive slime is an open-source LLM post-training framework for reinforcement learning scaling, connecting Megatron-based training with SGLang-base… | 79 | 8606 | active |
| bitsandbytes-foundation/bitsandbytes bitsandbytes is a Python library providing k-bit quantization primitives for PyTorch, enabling 8-bit (LLM.int8()) and 4-bit (QLoRA) quantiz… | 94 | 8514 | active |
| isaac-sim/IsaacLab Isaac Lab is a GPU-accelerated open-source framework for robot learning built on NVIDIA Isaac Sim, unifying workflows like reinforcement le… | 96 | 8397 | active |
| EricLBuehler/mistral.rs mistral.rs is a fast and flexible LLM inference engine written in Rust, supporting many model families with quantization (ISQ/UQFF, GGUF), … | 92 | 7741 | active |
| EleutherAI/gpt-neox GPT-NeoX is EleutherAI's library for training large-scale autoregressive transformer language models on GPUs, built on NVIDIA's Megatron an… | 67 | 7471 | active |
| naver/dust3r DUSt3R is the official PyTorch implementation of a CVPR 2024 model that performs dense, unconstrained stereo and multi-view 3D reconstructi… | 42 | 7317 | active |
| Paddle Lite Paddle Lite is a high-performance, lightweight deep learning inference engine from Baidu's PaddlePaddle ecosystem, designed for mobile, emb… | 55 | 7277 | active |
| kohya-ss/sd-scripts A collection of Python training, generation, and utility scripts for Stable Diffusion and other image generation models, most widely used f… | 96 | 7242 | active |
| BVLC/caffe Caffe is a fast, modular deep learning framework developed by Berkeley AI Research, focused on convolutional neural networks for vision, sp… | 23 | 34549 | maintenance |
| TheTom/turboquant_plus TurboQuant+ is a Python reference implementation of the TurboQuant KV cache compression method (ICLR 2026), using PolarQuant codebooks and … | 54 | 7036 | active |
| google-ai-edge/LiteRT-LM LiteRT-LM is Google's production-ready, high-performance open-source framework for running large language models on edge devices, built as … | 88 | 6702 | active |
| iperov/DeepFaceLive DeepFaceLive is a real-time face-swap application for PC streaming and video calls, using trained face models (DFM) applied to webcam or vi… | 10 | 31022 | maintenance |
| rtqichen/torchdiffeq torchdiffeq is a PyTorch library of differentiable ordinary differential equation (ODE) solvers, best known as the canonical implementation… | 35 | 6482 | stable |
| PufferLib PufferLib is a fast, open-source reinforcement learning library that trains tiny, super-human models in seconds, achieving 1M+ environment … | 87 | 6476 | active |
| ByteDance-Seed/Depth-Anything-3 Depth Anything 3 (DA3) is a transformer-based model that predicts spatially consistent depth and geometry from any number of visual inputs,… | 56 | 6445 | active |
| tensorflow/serving TensorFlow Serving is a flexible, high-performance serving system for machine learning models designed for production environments. It mana… | 89 | 6364 | stable |
| HarfBuzz HarfBuzz is a text shaping engine that converts Unicode text into properly positioned glyph output for any writing system, supporting OpenT… | 99 | 6165 | stable |
| volcano-sh/volcano Volcano is a CNCF-hosted, Kubernetes-native batch scheduling system that extends kube-scheduler for high-performance workloads like AI/ML t… | 98 | 5993 | stable |
| Michael-A-Kuykendall/shimmy Shimmy is a single-binary, OpenAI-compatible LLM inference server written in pure Rust, running GGUF models on a WebGPU-based engine (Airfr… | 86 | 5983 | active |
| fla-org/flash-linear-attention A PyTorch library providing hardware-efficient implementations of emerging sequence model architectures, including linear attention, sparse… | 86 | 5940 | active |
| GPUStack GPUStack is an open-source GPU cluster manager for AI model serving that orchestrates inference engines like vLLM, SGLang, and TensorRT-LLM… | 88 | 5844 | active |
| areal-project/AReaL AReaL is a large-scale asynchronous reinforcement learning system that bridges foundation model training with agent-based applications, sup… | 86 | 5842 | active |
| pytorch/torchtitan torchtitan is a PyTorch-native platform for large-scale training of generative AI models, offering a clean-room implementation of PyTorch's… | 81 | 5830 | active |
| Newton Newton is an open-source, GPU-accelerated physics simulation engine built on NVIDIA Warp, targeting roboticists and simulation researchers.… | 87 | 5813 | active |
| AidLux (AidLearning) AidLux (originally AidLearning) is an AIoT development platform that runs a native Ubuntu Linux environment with GUI, deep learning tooling… | 63 | 5812 | active |
| SmartSub SmartSub (妙幕) is a free, open-source cross-platform desktop application that provides an end-to-end subtitle and dubbing pipeline: speech-t… | 92 | 5776 | active |
| NVIDIA/DALI NVIDIA DALI is a GPU-accelerated data loading and preprocessing library with optimized building blocks and an execution engine for deep lea… | 91 | 5761 | active |
| Nuclio Nuclio is a high-performance open-source serverless (FaaS) platform for real-time event and data processing, deployable standalone via Dock… | 92 | 5760 | stable |
| pjreddie/darknet Darknet is an open-source neural network framework written in C and CUDA, best known as the original home of the YOLO real-time object dete… | 32 | 26503 | maintenance |
| nerfstudio gsplat is an open-source Python library with CUDA-accelerated, differentiable rasterization of Gaussians, based on 3D Gaussian Splatting fo… | 73 | 5716 | active |
| LaurentMazare/tch-rs tch-rs is a Rust crate providing thin bindings to the C++ API of PyTorch (libtorch), staying close to the original API. It enables tensor o… | 64 | 5503 | active |
| mosaicml/composer Composer is an open-source PyTorch-based deep learning training library by MosaicML (now Databricks) for training neural networks faster an… | 66 | 5499 | active |
| RLinf/RLinf RLinf is an open-source, flexible and scalable reinforcement learning training infrastructure for embodied AI (vision-language-action model… | 72 | 5453 | active |
| NVIDIA/Model-Optimizer NVIDIA Model Optimizer (ModelOpt) is a Python library of state-of-the-art model optimization techniques including quantization, pruning, di… | 92 | 5386 | active |
| beclab/Olares Olares is an open-source personal cloud operating system built on Kubernetes that turns your own hardware into a self-hosted AI platform fo… | 90 | 5299 | active |
| Transformer Lab Transformer Lab is an open-source desktop application (built with Electron and Python) that provides a unified GUI for training, fine-tunin… | 82 | 5195 | active |
| hiyouga/EasyR1 EasyR1 is an efficient, scalable reinforcement learning training framework for large language models and vision-language models, built as a… | 62 | 5192 | active |
| h2oai/h2o-llmstudio H2O LLM Studio is a framework and no-code GUI for fine-tuning state-of-the-art large language models, built by H2O.ai. It supports LoRA and… | 96 | 5184 | active |
| ArrayFire ArrayFire is a general-purpose tensor/numerical computing library for C, C++, and Python that accelerates array operations on GPUs (CUDA, O… | 65 | 4903 | stable |
| mindspore-ai/mindspore MindSpore is an open-source deep learning framework for training and inference across mobile, edge, and cloud scenarios. It provides automa… | 32 | 4707 | active |
| HAMi HAMi (Heterogeneous AI Computing Virtualization Middleware) is a CNCF Incubating, Kubernetes-native GPU virtualization and scheduling middl… | 91 | 4691 | active |
| Tencent/TNN TNN is a high-performance, lightweight deep learning inference framework developed by Tencent Youtu Lab, supporting mobile, desktop, and se… | 28 | 4656 | active |
| hao-ai-lab/FastVideo FastVideo is a unified Python framework for post-training and real-time inference of video diffusion models, covering data preprocessing, f… | 79 | 4616 | active |
| NVlabs/tiny-cuda-nn A small, self-contained C++/CUDA framework for training and querying neural networks, featuring a lightning-fast fully fused MLP and a vers… | 66 | 4536 | active |
| OAID/Tengine Tengine is a lightweight, high-performance, modular deep learning inference engine developed by OPEN AI LAB for embedded and edge devices. … | 24 | 4531 | active |
| iperov/DeepFaceLab DeepFaceLab is the leading open-source Windows application for creating deepfakes, allowing users to swap, de-age, or replace faces and hea… | 10 | 19274 | maintenance |
| ModelTC/LightLLM LightLLM is a Python-based LLM inference and serving framework designed for lightweight deployment, easy scalability, and high throughput. … | 82 | 4317 | active |
| isaac-sim/IsaacSim NVIDIA Isaac Sim is an open-source robotics simulation application built on NVIDIA Omniverse for developing, simulating, and testing AI-dri… | 83 | 4283 | active |
| zml/zml ZML is a production LLM inference stack written in Zig, built on MLIR and OpenXLA, that compiles models to run at peak performance across N… | 69 | 4136 | active |
| FedML FedML (TensorOpera) is a unified Python library for large-scale distributed training, model serving, and federated learning across GPU clou… | 41 | 4070 | active |
| IREE IREE is an MLIR-based end-to-end machine learning compiler and runtime that lowers models from frameworks like PyTorch, TensorFlow, JAX, an… | 95 | 3980 | active |
| Nunchaku Nunchaku is a high-performance inference engine for 4-bit quantized diffusion models (and LLMs) based on the SVDQuant technique from an ICL… | 73 | 3953 | active |
| thu-ml/TurboDiffusion TurboDiffusion is a Python framework that accelerates end-to-end video diffusion model generation by 100-200x using SageAttention, Sparse-L… | 58 | 3851 | active |
| MrNeRF/LichtFeld-Studio LichtFeld Studio is a native open-source desktop application for 3D Gaussian Splatting that combines training, real-time inspection, splat … | 96 | 3849 | active |
| kaldi-asr/kaldi Kaldi is a C++ toolkit for speech recognition research and development, including acoustic modeling, feature extraction, decoding, and spea… | 48 | 15502 | maintenance |
| xororz/local-dream A free, open-source Android app for running Stable Diffusion locally with Snapdragon NPU acceleration, also supporting CPU/GPU inference. I… | 82 | 3598 | active |
| NVIDIA/TransformerEngine Transformer Engine is an NVIDIA library for accelerating Transformer model training and inference on NVIDIA GPUs using low-precision format… | 99 | 3576 | active |
| ashvardanian/StringZilla StringZilla is a high-performance string processing library for C, C++, Python, Rust, Swift, JS, and Go that uses SIMD, SWAR, and GPU instr… | 94 | 3570 | active |
| guandeh17/Self-Forcing Official implementation of Self Forcing, a training method for autoregressive video diffusion models that simulates inference during traini… | 34 | 3551 | active |
| Mesh-LLM/mesh-llm Mesh LLM is a Rust-based distributed LLM inference runtime that pools GPUs and memory across machines into a single OpenAI-compatible API. … | 79 | 3522 | active |
| huggingface/optimum Optimum is a Hugging Face library that extends Transformers, Diffusers, timm, and Sentence Transformers with hardware-specific optimization… | 91 | 3503 | active |
| alibaba/ROLL ROLL is an open-source reinforcement learning library from Alibaba for training large language models at scale, supporting algorithms like … | 82 | 3429 | active |
| superlinked/sie SIE (Superlinked Inference Engine) is an open-source, self-hosted inference server and production cluster that serves 100+ open models (emb… | 95 | 3414 | active |
| PhysicsNeMo NVIDIA PhysicsNeMo is an open-source Python deep-learning framework for building, training, fine-tuning, and inferring physics AI models us… | 93 | 3373 | active |
| LeelaChessZero/lc0 Lc0 is an open-source, UCI-compliant chess engine that plays chess using neural networks trained via AlphaZero-style self-play reinforcemen… | 74 | 3248 | active |
| Jittor/jittor Jittor is a high-performance deep learning framework from Tsinghua University based on just-in-time (JIT) compilation and meta-operators, w… | 68 | 3237 | active |
| Arm NN Arm's Compute Library is a C++ collection of over 100 low-level machine learning and computer vision functions optimized for Arm Cortex-A/N… | 98 | 3197 | active |
| Miles Miles is an open-source, enterprise-grade reinforcement learning framework for large-scale LLM and VLM post-training, forked from and co-ev… | 84 | 3169 | active |
| mjlab mjlab is a Python framework that combines Isaac Lab's manager-based API with MuJoCo Warp, a GPU-accelerated version of MuJoCo, for reinforc… | 82 | 3163 | active |
| luminal-ai/luminal Luminal is a high-performance general-purpose ML inference compiler written in Rust that lowers models to a minimal 15-op dataflow IR and c… | 77 | 3026 | active |
| pytorch/TensorRT Torch-TensorRT is a compiler library that accelerates PyTorch model inference on NVIDIA GPUs using TensorRT. It supports just-in-time compi… | 93 | 3009 | active |
| pytorch/ao TorchAO is a PyTorch-native library for model optimization through quantization and sparsity. It supports quantizing weights, gradients, op… | 87 | 3001 | active |
| bghira/SimpleTuner SimpleTuner is a Python fine-tuning toolkit for image, video, and audio diffusion models built on Hugging Face Diffusers. It provides a web… | 93 | 2940 | active |
| Mitsuba 3 Mitsuba 3 is a research-oriented, retargetable rendering system for forward and inverse light transport simulation, written in C++17 on top… | 90 | 2928 | active |
| ModelTC/LightX2V LightX2V is a lightweight, high-performance inference framework for image and video generation, supporting tasks like text-to-video, image-… | 84 | 2914 | active |
| antirez/h3.c A native C inference engine for the MiniMax H3 model on Apple Silicon, using Metal for GPU acceleration. It generates video (and audio) fro… | 53 | 2828 | active |
| pytorch/xla PyTorch/XLA is a Python package that connects the PyTorch deep learning framework to XLA devices such as Google Cloud TPUs via the XLA deep… | 63 | 2801 | active |
| NVlabs/stylegan2 The official TensorFlow implementation of StyleGAN2, NVIDIA's improved style-based generative adversarial network for high-quality uncondit… | 32 | 11180 | maintenance |
| huggingface/text-generation-inference Text Generation Inference (TGI) is a Rust, Python and gRPC toolkit for deploying and serving large language models with high performance, p… | 10 | 10877 | maintenance |
| openlake-project/openlake OpenLake is a high-performance distributed storage engine written in Rust (built on io_uring, RDMA, and GPUDirect) designed to feed GPUs du… | 78 | 2662 | active |
| NVlabs/LongLive LongLive is an NVIDIA research framework providing parallel training and inference infrastructure for real-time long video generation, usin… | 60 | 2639 | active |