A curated directory of high-quality, truly open-source AI projects, models, tools, and infrastructure, this repository is designed for technical users to rapidly assess and select AI components. It covers the full spectrum from core frameworks and foundation models to inference, deployment, MLOps, and specialized applications, facilitating informed adoption decisions.

How It Works

The list is meticulously organized into 14 detailed categories, including core libraries, open foundation models, inference engines, agentic systems, generative media tools, and MLOps platforms. Each entry highlights key projects, their primary use cases, and relevant links, facilitating efficient discovery and evaluation of open-source AI solutions.

Quick Start & Requirements

This repository is a curated list, not a runnable project. No installation is required; it serves as a discovery index for other open-source projects.

Highlighted Details

• Core Frameworks & Libraries: Features leading deep learning frameworks like PyTorch, TensorFlow, and JAX, alongside essential NLP libraries such as Hugging Face Transformers, and data manipulation tools like Pandas and Polars.
• Open Foundation Models: Covers recent open-weight models including Qwen, Gemma, Mistral, Phi, and LLaVA, spanning LLMs, coding models, multimodal capabilities, and generative media.
• Inference & Serving: Includes efficient local inference engines like llama.cpp and Ollama, and high-performance production servers such as vLLM and Text Generation Inference (TGI).
• Agentic AI & RAG: Details frameworks for building agents (AutoGen, CrewAI, LangGraph) and essential Retrieval-Augmented Generation (RAG) components like vector databases (Chroma, Qdrant) and retrieval frameworks (LlamaIndex, Haystack).
• Generative Media: Showcases state-of-the-art open models and tools for image (Stable Diffusion, ComfyUI, Fooocus), video (Open-Sora), and audio generation.
• MLOps & Evaluation: Lists essential tooling for production AI, such as MLflow, ClearML, Langfuse for observability, and benchmark suites like lm-evaluation-harness and HELM.

Maintenance & Community

Projects included are vetted for active maintenance, requiring commits within the last six months. Contributions are welcomed via Pull Requests, with detailed guidelines provided in the CONTRIBUTING.md file.

Licensing & Compatibility

The curated list itself is licensed under CC0 1.0 Universal, allowing for unrestricted use. However, the individual open-source projects listed are subject to their own licenses, which may vary. Users must independently verify compatibility for commercial or closed-source applications.

Limitations & Caveats

This resource's primary limitation is its reliance on the accuracy and up-to-dateness of the included project links and descriptions. The "truly open-source" status is determined by the curator and requires independent verification by the user. The rapid pace of AI development necessitates continuous updates to maintain comprehensiveness.