Projects

Overview: This repository contains a comprehensive list of 500 AI, machine learning, deep learning, computer vision, and NLP projects with code, designed to help developers and learners enhance their skills in these areas. Key features: - 500+ projects across various domains - Projects are linked to their respective GitHub repositories - All projects are tested and working fine - Contributors are encouraged to contribute and ping if any links don't work Tech stack: Python, JavaScript, HTML/CSS, Markdown, and various libraries and frameworks Why it matters: This repository serves as a valuable resource for developers and learners looking to expand their knowledge in AI, machine learning, deep learning, computer vision, and NLP. It provides a wide range of projects that can be used for educational purposes or practical applications.

### Overview: SGLang is a high-performance inference framework designed for large language models and multimodal models. Optimized for agentic workloads, SGLang is built to serve models efficiently and reliably, making it ideal for large-scale applications and real-time processing. ### Key Features: - **SGLang Diffusion**: Built into the framework, SGLang Diffusion is a built-in image and video generation engine for diffusion models. - **Model Compatibility**: Supports a wide range of GPUs, TPUs, NPUs, CPUs, and Apple Silicon platforms. - **Documentation and Community**: Includes comprehensive documentation, a cookbook, and community resources such as meetups, developer meetings, workshops, and office hours. ### Tech Stack: Python ### Why It Matters: SGLang is crucial for developers looking to optimize and scale their large language models and multimodal applications. Its performance and compatibility with various hardware platforms make it an essential tool for modern AI development and deployment. Repository: sgl-project/sglang GitHub Stars: 37,258

Overview: This project is an AI-powered website cloning tool that allows developers to clone any website with a single command using their AI coding agents. Key features: - Clone any website - Use Claude Code or Opus 5.5 for best results - Supports Codex, Cursor, and OpenCode - GitHub stars: 36,937 Tech stack: TypeScript Why it matters: This tool simplifies the process of creating clean Next.js apps from any website, making it easier for developers to work with and maintain their projects.

### Overview: MockingBird is a real-time voice cloning tool that generates arbitrary speech in 5 seconds using PyTorch and WebRTC. ### Key features: - 🚧 Supports Chinese (Mandarin) and multiple datasets like aidatatang_200zh, magicdata, aishell3, and more. - 🚀 Works with PyTorch 1.9.0 and GPU Tesla T4 and GTX 2060. - 🚀 Runs on Windows, Linux, and M1 Macs. - 🚀 Easy-to-use effect with a newly-trained synthesizer. - 🚀 Ready to serve results via a web server. ### Tech stack: Python, PyTorch, WebRTC, WebSockets, Webpack, and Webpack Dev Server. ### Why it matters: MockingBird simplifies the process of voice cloning, making it accessible and efficient for developers and creators.

Overview: FreeLLMAPI aggregates free tiers from dozens of LLM providers, offering a combined capacity of 7.4 billion tokens per month across 635 free model endpoints. It provides a single OpenAI-compatible endpoint for easy integration and usage. Key features: - Aggregates free tiers from 34 providers - Offers a unified /v1 API endpoint - Supports automatic failover and smart routing Tech stack: TypeScript Why it matters: FreeLLMAPI simplifies access to a vast pool of free LLM resources, enabling developers to experiment and build applications without the need for costly subscriptions.

**Overview:** HanLP is a comprehensive natural language processing (NLP) toolkit designed for both research and production environments. It supports multiple languages and tasks, including Chinese part-of-speech tagging, named entity recognition, dependency parsing, and semantic role labeling, among others. **Key features:** - **Part-of-Speech Tagging (POS):** Supports multiple languages and tasks. - **Named Entity Recognition (NER):** Recognizes named entities in text. - **Dependency Parsing (DP):** Analyzes the syntactic structure of sentences. - **Semantic Role Labeling (SRL):** Identifies the roles of words in sentences. - **Sentence Similarity (SS):** Measures the similarity between sentences. - **New Word Discovery (NWD):** Finds new words and their meanings. - **Keyword Extraction (KE):** Extracts key phrases and words. - **Automatic Summarization (AS):** Generates summaries of text. - **Text Classification (TC):** Classifies text into predefined categories. - **Text Clustering (TC):** Groups similar text segments. - **Pinyin to Simplified Chinese Conversion (P2SC):** Converts pinyin to simplified Chinese characters. - **Automatic Summary (AS):** Generates summaries of text. - **Text Classification (TC):** Classifies text into predefined categories. - **Text Clustering (TC):** Groups similar text segments. - **Pinyin to Simplified Chinese Conversion (P2SC):** Converts pinyin to simplified Chinese characters. **Tech stack:** HanLP is built on PyT

Overview: GraphRAG is a modular graph-based Retrieval-Augmented Generation (RAG) system designed to enhance LLMs with targeted context from narrative data. This project explores the integration of graphs to unlock LLMs' capabilities for private data discovery. Key features: - Graph-based retrieval to provide context for LLM outputs - Integration with LLMs to enhance reasoning abilities - Support for private data analysis Tech stack: Python Why it matters: GraphRAG leverages the power of graphs to enhance the functionality of LLMs, making it a valuable tool for researchers and developers looking to improve the reasoning capabilities of their models on sensitive data.

Overview: Next AI Draw.io allows users to draw and edit Draw.io diagrams by chatting with an AI, creating architecture diagrams, flowcharts, sequence diagrams, and more from one sentence. It supports various models and provides versions, undo, and export options. Key features: - Draw architecture diagrams, flowcharts, sequence diagrams, and more from one sentence, with built-in AWS, Azure, GCP, and Kubernetes icon libraries. - Connectors can carry a flowing animation. - Upload a screenshot or sketch and the AI redraws it; upload PDF, Markdown, code, and other text files and get a diagram of their content. - Edit by chat: changes stream onto the canvas, and the AI highlights the shapes it just changed. - Versions and undo: every AI change is a version card with a thumbnail; compare it with the canvas, restore it, or undo it, and Ctrl+Z on the canvas also takes back an AI change in one step. - It is a normal Draw.io diagram: double-click to rename, drag, restyle, use several pages, and export as `.drawio`, `.png`, `.svg`, or `.drawio.svg` at any time. Tech stack: TypeScript Why it matters: Next AI Draw.io combines the power of AI with the flexibility of Draw.io, making it a versatile tool for anyone looking to create diagrams quickly and easily.

### Overview: AgentGPT is a powerful tool that allows you to configure and deploy autonomous AI agents in your browser. It simplifies the process of assembling, configuring, and deploying AI agents, making it accessible and easy to use. ### Key Features: - 🚀 Assemble and configure AI agents for any task. - 📂 Deploy agents with a database and backend. - 🧑‍💻 Frontend for a seamless user experience. - 🚀 Execute tasks and learn from results. ### Tech Stack: - ✅ TypeScript - ✅ FastAPI - ✅ Nextjs - ✅ Docker - ✅ OpenAI API key - ✅ Serper API Key (optional) - ✅ Replicate API Token (optional) ### Why It Matters: AgentGPT simplifies the process of deploying AI agents, making it accessible and easy to use. Whether you're a developer or a business owner, AgentGPT can help automate tasks and make your life easier.

Overview: Langfuse is an open-source platform designed to help teams develop, monitor, evaluate, and debug AI applications collaboratively. It is battle-tested and can be self-hosted in minutes. Key features: - Collaborative development and debugging - Real-time monitoring and evaluation - Traceability and observability - Self-hosting capability Tech stack: TypeScript Why it matters: Langfuse provides a comprehensive solution for developers and engineers looking to improve the quality and performance of their AI applications. Its open-source nature and battle-tested self-hosting capability make it an attractive choice for teams seeking to enhance their AI development workflow.

### Overview: This curated collection of Agent Skills is a repository of real-world Agent Skills created and used by actual engineering teams, not AI-sloped generated content. It includes skills compatible with Claude Code, Codex, Antigravity, Gemini CLI, Cursor, GitHub Copilot, OpenCode, Windsurf, and more. ### Key features: - Compatible with Claude Code, Codex, Antigravity, Gemini CLI, Cursor, GitHub Copilot, OpenCode, Windsurf, and more. - Skills curated by leading development teams and the community. - Skills compatible with various AI and development tools. - Skills are hand-picked, not AI-sloped generated. ### Tech stack: Officially maintained by the VoltAgent team. ### Why it matters: This collection provides a valuable resource for developers looking to integrate AI and automation into their projects, ensuring that skills are vetted and used effectively.

Overview: DeepSeek-Reasonix is a reliable coding agent designed for complex software engineering tasks, providing a model of your choice to run on your computer. Key features: - Model of choice - Supports multiple programming languages - Intuitive user interface - Advanced debugging tools - Extensive documentation and community support Tech stack: Go Why it matters: DeepSeek-Reasonix offers a powerful and efficient solution for software developers, enhancing productivity and project quality through its model-based approach and comprehensive features. Repository: esengine/DeepSeek-Reasonix

Overview: QwenPaw is a powerful AI assistant that can be deployed locally or on the cloud, providing easy installation and support for multiple chat apps. It offers extensible capabilities and integrates seamlessly with various channels. Key features: - Deployed locally or in the cloud - Supports multiple chat apps - Extensible capabilities - Easy installation Tech stack: TypeScript Why it matters: QwenPaw is designed to be a versatile and flexible AI assistant, making it suitable for various deployment environments and integration with different platforms. Repository: agentscope-ai/QwenPaw GitHub stars: 35,809 GitHub description: Your Personal AI Assistant; easy to install, deploy on your own machine or on the cloud; supports multiple chat apps with easily extensible capabilities.

### Overview: The Oh My Pi (OMP) coding agent is a highly capable and continuously tuned IDE built by Stencil Labs. It integrates seamlessly with TypeScript, Rust, and other popular programming languages, providing a robust and user-friendly development environment. ### Key Features: - **TypeScript Support**: Seamless integration with TypeScript for enhanced productivity. - **Rust Core**: Built with Rust, ensuring performance and reliability. - **LSP Operations**: Advanced Language Server Protocol (LSP) operations for efficient code navigation and debugging. - **DAP Operations**: Debugging and profiling capabilities with the Debugger Adapter Protocol (DAP). - **Built-in Tools**: A wide range of tools for common programming tasks, including linters, formatters, and more. - **Continuous Improvement**: Continuously tuned by real-world use, ensuring optimal performance and reliability. ### Tech Stack: TypeScript ### Why It Matters: OMP is the most capable agent surface that ships, continuously tuned by real-world use, complete out of the box, and open all the way down. It integrates seamlessly with TypeScript, Rust, and other popular programming languages, providing a robust and user-friendly development environment.

### Overview: Vibe-Trading is a comprehensive trading agent designed to empower users with advanced trading capabilities. With its intuitive interface and robust features, it simplifies the process of trading and analysis. ### Key features: - **Real-time Trading**: Access to live market data and real-time trading capabilities. - **Advanced Analysis**: Comprehensive financial analysis tools for in-depth market insights. - **Customizable Alerts**: Set up alerts for specific market conditions or price movements. - **Shadow Account**: A secure and private trading account for enhanced privacy. ### Tech stack: - **Python**: The primary programming language used for the backend and frontend development. - **FastAPI**: For the backend API, providing a fast and efficient server. - **React**: For the frontend, offering a user-friendly and responsive interface. ### Why it matters: Vibe-Trading is designed to be a reliable and efficient trading tool, making it ideal for both beginners and experienced traders. Its advanced analysis features and secure shadow account options make it a valuable asset in the trading world. ### Repository: - **Repository**: HKUDS/Vibe-Trading - **Stars**: 35,728 - **Description**: "Vibe-Trading: Your Personal Trading Agent" ### Tech Stack: - **Python**: The primary language used for the backend and frontend development. - **FastAPI**: For the backend API, providing a fast and efficient server. - **React**: For the frontend, offering a user-friendly and responsive interface. ### Why it matters: Vibe-Trading is designed to be a reliable and efficient

### Overview: DragGAN is an innovative generative model designed for interactive manipulation of generative image manifolds. This project showcases the ability to manipulate images in real-time, making it highly valuable for applications requiring dynamic and responsive visual effects. ### Key Features: - **Interactive Manipulation**: Users can drag and manipulate images on the generative image manifold in real-time. - **Dynamic Visual Effects**: The model allows for the creation of dynamic and interactive visual effects, enhancing the user experience. - **Versatile Applications**: DragGAN can be applied across various domains such as art, design, and interactive storytelling. ### Tech Stack: - Python ### Why It Matters: DragGAN offers a significant advancement in the field of generative models by providing interactive manipulation capabilities, making it highly valuable for applications requiring dynamic visual effects. Its real-time manipulation feature sets it apart, making it a powerful tool for creators and developers alike. ### Repository: XingangPan/DragGAN - **Primary Language**: Python - **GitHub Stars**: 35,728 - **GitHub Description**: Official Code for DragGAN (SIGGRAPH 2023) - **README Excerpt**: Official Code for DragGAN (SIGGRAPH 2023). The project showcases the ability to manipulate images on the generative image manifold in real-time. It is part of the SIGGRAPH 2023 conference proceedings. ### Web Demos: - **OpenXLab App**: [OpenXLab](https://openxlab.org.cn/apps/detail/XingangPan/DragGAN

### Overview: Heretic is a fully automatic tool designed to remove censorship from transformer-based language models without the need for expensive post-training. It combines advanced directional ablation techniques with a TPE-based parameter optimizer, enabling it to find high-quality abliteration parameters automatically. ### Key Features: - **Automatic Abliteration**: Heretic automatically removes censorship from language models without requiring manual intervention. - **Comprehensive Support**: Supports most dense models, including multimodal and hybrid models like Qwen3.5. - **Efficient Training**: Works completely automatically, making it accessible to anyone familiar with command-line programming. ### Tech Stack: - Python ### Why It Matters: Heretic simplifies the process of removing censorship from language models, making it easier for developers and researchers to experiment with and improve their models without the need for advanced knowledge of transformer internals. This tool ensures that decensored models retain as much of the original model's intelligence as possible, enhancing the overall quality of the language generation.

### Overview: Turn any technical book, document folder, or collection of sources into a unified agent skill — ready to study, reference, and use while you work in GitHub Copilot CLI, Amp, Claude Code, Codex, Hermes Agent, OpenCode, or OpenClaw. ### Key features: - **Skill Generation**: Convert books, documents, and collections into agent skills. - **Versatile Formats**: Supports PDF, EPUB, DOCX, MD, HTML, RTF, and MOBI formats. - **Real-World Integration**: Compatible with popular AI tools like GitHub Copilot CLI, Amp, Claude Code, Codex, Hermes Agent, OpenCode, and OpenClaw. ### Tech stack: - Python ### Why it matters: - **Efficient Study**: Quickly access the knowledge from a book without the need to read it again. - **Contextual Understanding**: Provides insights and decision rules directly from the book, enhancing comprehension. - **Versatile Use**: Easily integrate into various AI tools for enhanced productivity and research. ### GitHub stars: 351,600

### Overview: Qdrant is a high-performance, massive-scale vector database and search engine designed for the next generation of AI applications. It provides a production-ready service with a convenient API for storing, searching, and managing points—vectors with an additional payload. Qdrant is written in Rust, making it fast and reliable even under high load. It offers extended filtering support, suitable for applications like matching, searching, and recommending. ### Key features: - **Vector Storage**: Stores and indexes vectors efficiently. - **Querying**: Supports vector similarity search and retrieval. - **Extended Filtering**: Enables advanced filtering capabilities. - **Cloud Integration**: Available as a fully managed service in the cloud. ### Tech stack: Rust ### Why it matters: Qdrant is a crucial tool for developers looking to integrate vector search capabilities into their AI applications, making it easier to build matching, searching, and recommendation systems. Its fast performance and reliability make it an essential component in modern AI development. ### Repository: Repository: qdrant/qdrant GitHub stars: 35,118

Overview: This repository is a collection of open-source implementations of various algorithms in C++, licensed under the MIT License, aimed at providing educational resources for students and educators. Key features: - Atomic source codes using STL classes without external libraries. - Well-documented source code with detailed explanations. - Compiles and tests on Windows, MacOS, and Ubuntu. - Adheres to C++17 standards for portability. - Ensures correct implementations with self-checks. - Modular and open-source licensing. Tech stack: C++ Why it matters: This repository offers a comprehensive and educational resource for learning and implementing algorithms in C++, making it valuable for both students and educators in various fields such as computer science, mathematics, and data science.