Fastest-Growing AI Open Source Projects by Stars in 2025

An open-source Faircode workflow automation platform with native AI capabilities, combining visual building and custom code, supporting 400+ integrations.

A powerful low-code AI application building platform designed for creating intelligent agents and RAG applications.

Serena is a powerful coding agent toolkit that transforms large language models (LLMs) into a fully functional agent, working directly on your codebase.

An open-source Retrieval-Augmented Generation engine based on deep document understanding, providing accurate and reliable question answering capabilities for businesses of all sizes.

A feature-rich, self-hosted AI chat interface that supports Ollama and OpenAI API, providing a completely offline AI conversation experience.

vllm is a fast and easy-to-use library for fast inference of large language models.

An open-source LLM application development platform that integrates agent workflows, RAG pipelines, model management, and more, allowing you to quickly go from prototype to production deployment.

Ollama: Run, create, and share large language models locally.

Playwright-based Model Context Protocol (MCP) server providing browser automation capabilities for LLMs

The easiest way for AI agents to connect to and control browsers, enabling website automation.

GitHub MCP Server is a Model Context Protocol (MCP) server designed to seamlessly integrate with the GitHub API, providing developers and tools with advanced automation and interaction capabilities.

AI Query Engine - An open-source platform to build and deploy machine learning models on large-scale federated data sources using SQL syntax.

Unsloth: Efficient fine-tuning and reinforcement learning tool for large language models - 2x faster training, 70% less memory usage

A lightweight multi-agent system development framework with support for memory, knowledge, and reasoning capabilities.

An open-source document intelligent Q&A tool based on RAG technology, supporting conversational interaction with your documents.

A lightweight multi-agent workflow building framework that supports inter-agent collaboration, tool invocation, and process orchestration.

A no-code, low-code large language model fine-tuning and deployment framework that supports unified and efficient fine-tuning of 100+ LLMs/VLMs.

An open-source drag-and-drop visual AI agent building platform based on LangChain technology, allowing users to quickly create custom LLM applications without programming.

Open-source, high-performance web crawler and data extraction tool optimized for LLMs and AI agents.

An AI large model-based one-click HD short video generation tool

A multi-agent collaborative automation framework that enables AI agents to work together like team members to execute complex tasks.

A real-time temporal knowledge graph construction framework for AI agents

An open-source multi-agent AI application framework developed by Microsoft, supporting multiple AI agents to collaborate on solving complex tasks.

A Gradio-based Stable Diffusion Web UI interface, providing AI painting functions such as text-to-image and image-to-image.

Transformers provides thousands of pre-trained models to support tasks in over 100 languages across text, vision, audio, and more.

Open-source, all-in-one AI desktop application with built-in RAG, AI agents, and a no-code agent builder.

An open-source framework for building MCP servers and clients in a fast, Pythonic way.

llama.cpp is a LLaMA model inference engine written in pure C/C++, designed for high performance and low resource usage.

GPT-SoVITS: A few-shot voice cloning tool that can train a high-quality TTS model with just 1 minute of voice data.

An AI assistant development framework for React applications, quickly build production-grade AI chatbots and intelligent agents

PaddleOCR aims to create a rich, leading, and easy-to-use OCR toolkit to help developers train better, faster, and more accurate OCR models.

Unified LLM API gateway, supporting OpenAI format calls for 100+ large language model providers

An open-source, modern AI chat framework that supports multiple AI providers, knowledge base management, multimodal interaction, and more.

PyTorch: Python tensor and dynamic neural network with strong GPU acceleration

CosyVoice: A multilingual large-scale speech generation model providing full-stack capabilities for inference, training, and deployment.

A TypeScript-based AI agent framework for quickly building AI applications and features.

Coqui TTS: A deep learning toolkit for text-to-speech, proven through research and production practice.

LightRAG is a simple and fast Retrieval-Augmented Generation framework that supports multiple query modes and knowledge graph construction.

Oneflow is a deep learning framework designed for production environments. It features automatic parallelism and distributed training.

A fast high-level Python web crawling and web scraping framework used to crawl websites and extract structured data from pages.

A high-performance C/C++ port of the OpenAI Whisper speech recognition model, supporting pure CPU inference and multi-platform deployment.

Dia: A text-to-speech (TTS) model capable of generating hyper-realistic conversations in one go.

An efficient and lightweight deep learning framework optimized for mobile and embedded devices, supporting model inference and training.

SOTA open-source Text-to-Speech (TTS) system

Exo: Run your own AI cluster at home using everyday devices

Desktop Commander MCP is an MCP server designed for Claude, giving it terminal control, file system search, and diff file editing capabilities.

A knowledge graph-based retrieval-augmented generation system that automatically extracts structured knowledge graphs from text using LLMs and enhances RAG performance.

Streamlit is a Python library that allows you to create beautiful, custom web applications with pure Python code for machine learning, data science, and more.

A browser automation tool based on the MCP protocol, allowing AI applications to control your browser.

A leading data framework for building LLM-powered intelligent agents, specializing in private data augmentation.