The database of free AI tools on GitHub

Run LLMs like LLaMA on local devices with C#/.NET.

Automates incident investigation and root cause analysis with AI.

Optimize and accelerate PyTorch models for inference and training.

Self-host a complete LLM stack with Harbor CLI.

Accelerate LLM inference by clustering home devices.

Simplifies local AI model serving and inference through containerization.

A metasearch library that aggregates results from various web search services.

高性能大模型推理框架，支持多种算力和场景。

Serves text-embeddings and reranking models with low latency.

Self-hosted file-processing and AI tool for images, videos, audio, PDFs, and documents.

Compresses large language models for efficient inference.

Rust library for fast ONNX model inference and training.

Orchestrate data, pipelines, and models for AI applications.

Run AI models locally with node.js bindings for llama.cpp.

Lightweight ONNX inference library for AI models on resource-constrained devices.

🔥 Real-time NVIDIA GPU dashboard

Real-time 3D full-body reconstruction from a single camera, Multiperson BVH output, Pure C++ runtime, ONNX + ggml, 70-joint skeleton with hands.

eLLM: Run Long-Horizon Inference Faster on CPUs Than on GPUs

Helm chart for Ollama on Kubernetes

a community oriented 1:1, vLLM-alike (Continuous batching, paged KV) engine in C++ with additional features (GGUF, RadixAttention, Cache-aware scheduling, ...)

search every screenshot you've ever taken by the text inside it. native linux, offline, PaddleOCR + sqlite + gpui

Pure-Rust, CPU-only OCR engine for Baidu Unlimited-OCR (a DeepSeek-OCR-derived 3B MoE VLM). Five-model zoo, custom int8 kernels, no ML framework, no Python, no GPU.

Official Hi3D v1.0: high-resolution 3D geometry refinement built on Sparc3D. Inference code, model weights, and documentation.

AMD MIVisionX is a computer vision toolkit built around a highly optimized, conformant open-source implementation of the Khronos OpenVX™ 1.3.2 specification. As of the 4.0.0 release, MIVisionX ships t

Snap any image, screenshot, or webpage into plaintext. No GPU. No cloud. One command.

Frontier: A Discrete-Event Simulator for Modern LLM Serving

llama.cpp fork that abliterates at inference

CPU-first LLM inference server on llama.cpp. Runs useful models on free-tier ARM boxes; rewriting the input made it ~6x faster and more accurate than tuning the engine. MIT, benchmarks and failures in

✈️ Kubernetes-native platform for deploying and managing AI inference across multiple providers

hummingbird is a lightweight, zero dependency runtime for massive open source Mixture of Experts (MoE) language models. It unifies SSD, RAM, and VRAM into a single intelligent memory hierarchy, enabli