[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"similar-Andrew-Jang--RAGHub":3,"tool-Andrew-Jang--RAGHub":64},[4,17,27,35,43,56],{"id":5,"name":6,"github_repo":7,"description_zh":8,"stars":9,"difficulty_score":10,"last_commit_at":11,"category_tags":12,"status":16},3808,"stable-diffusion-webui","AUTOMATIC1111\u002Fstable-diffusion-webui","stable-diffusion-webui 是一个基于 Gradio 构建的网页版操作界面，旨在让用户能够轻松地在本地运行和使用强大的 Stable Diffusion 图像生成模型。它解决了原始模型依赖命令行、操作门槛高且功能分散的痛点，将复杂的 AI 绘图流程整合进一个直观易用的图形化平台。\n\n无论是希望快速上手的普通创作者、需要精细控制画面细节的设计师，还是想要深入探索模型潜力的开发者与研究人员，都能从中获益。其核心亮点在于极高的功能丰富度：不仅支持文生图、图生图、局部重绘（Inpainting）和外绘（Outpainting）等基础模式，还独创了注意力机制调整、提示词矩阵、负向提示词以及“高清修复”等高级功能。此外，它内置了 GFPGAN 和 CodeFormer 等人脸修复工具，支持多种神经网络放大算法，并允许用户通过插件系统无限扩展能力。即使是显存有限的设备，stable-diffusion-webui 也提供了相应的优化选项，让高质量的 AI 艺术创作变得触手可及。",162132,3,"2026-04-05T11:01:52",[13,14,15],"开发框架","图像","Agent","ready",{"id":18,"name":19,"github_repo":20,"description_zh":21,"stars":22,"difficulty_score":23,"last_commit_at":24,"category_tags":25,"status":16},1381,"everything-claude-code","affaan-m\u002Feverything-claude-code","everything-claude-code 是一套专为 AI 编程助手（如 Claude Code、Codex、Cursor 等）打造的高性能优化系统。它不仅仅是一组配置文件，而是一个经过长期实战打磨的完整框架，旨在解决 AI 代理在实际开发中面临的效率低下、记忆丢失、安全隐患及缺乏持续学习能力等核心痛点。\n\n通过引入技能模块化、直觉增强、记忆持久化机制以及内置的安全扫描功能，everything-claude-code 能显著提升 AI 在复杂任务中的表现，帮助开发者构建更稳定、更智能的生产级 AI 代理。其独特的“研究优先”开发理念和针对 Token 消耗的优化策略，使得模型响应更快、成本更低，同时有效防御潜在的攻击向量。\n\n这套工具特别适合软件开发者、AI 研究人员以及希望深度定制 AI 工作流的技术团队使用。无论您是在构建大型代码库，还是需要 AI 协助进行安全审计与自动化测试，everything-claude-code 都能提供强大的底层支持。作为一个曾荣获 Anthropic 黑客大奖的开源项目，它融合了多语言支持与丰富的实战钩子（hooks），让 AI 真正成长为懂上",138956,2,"2026-04-05T11:33:21",[13,15,26],"语言模型",{"id":28,"name":29,"github_repo":30,"description_zh":31,"stars":32,"difficulty_score":23,"last_commit_at":33,"category_tags":34,"status":16},2271,"ComfyUI","Comfy-Org\u002FComfyUI","ComfyUI 是一款功能强大且高度模块化的视觉 AI 引擎，专为设计和执行复杂的 Stable Diffusion 图像生成流程而打造。它摒弃了传统的代码编写模式，采用直观的节点式流程图界面，让用户通过连接不同的功能模块即可构建个性化的生成管线。\n\n这一设计巧妙解决了高级 AI 绘图工作流配置复杂、灵活性不足的痛点。用户无需具备编程背景，也能自由组合模型、调整参数并实时预览效果，轻松实现从基础文生图到多步骤高清修复等各类复杂任务。ComfyUI 拥有极佳的兼容性，不仅支持 Windows、macOS 和 Linux 全平台，还广泛适配 NVIDIA、AMD、Intel 及苹果 Silicon 等多种硬件架构，并率先支持 SDXL、Flux、SD3 等前沿模型。\n\n无论是希望深入探索算法潜力的研究人员和开发者，还是追求极致创作自由度的设计师与资深 AI 绘画爱好者，ComfyUI 都能提供强大的支持。其独特的模块化架构允许社区不断扩展新功能，使其成为当前最灵活、生态最丰富的开源扩散模型工具之一，帮助用户将创意高效转化为现实。",107662,"2026-04-03T11:11:01",[13,14,15],{"id":36,"name":37,"github_repo":38,"description_zh":39,"stars":40,"difficulty_score":23,"last_commit_at":41,"category_tags":42,"status":16},3704,"NextChat","ChatGPTNextWeb\u002FNextChat","NextChat 是一款轻量且极速的 AI 助手，旨在为用户提供流畅、跨平台的大模型交互体验。它完美解决了用户在多设备间切换时难以保持对话连续性，以及面对众多 AI 模型不知如何统一管理的痛点。无论是日常办公、学习辅助还是创意激发，NextChat 都能让用户随时随地通过网页、iOS、Android、Windows、MacOS 或 Linux 端无缝接入智能服务。\n\n这款工具非常适合普通用户、学生、职场人士以及需要私有化部署的企业团队使用。对于开发者而言，它也提供了便捷的自托管方案，支持一键部署到 Vercel 或 Zeabur 等平台。\n\nNextChat 的核心亮点在于其广泛的模型兼容性，原生支持 Claude、DeepSeek、GPT-4 及 Gemini Pro 等主流大模型，让用户在一个界面即可自由切换不同 AI 能力。此外，它还率先支持 MCP（Model Context Protocol）协议，增强了上下文处理能力。针对企业用户，NextChat 提供专业版解决方案，具备品牌定制、细粒度权限控制、内部知识库整合及安全审计等功能，满足公司对数据隐私和个性化管理的高标准要求。",87618,"2026-04-05T07:20:52",[13,26],{"id":44,"name":45,"github_repo":46,"description_zh":47,"stars":48,"difficulty_score":23,"last_commit_at":49,"category_tags":50,"status":16},2268,"ML-For-Beginners","microsoft\u002FML-For-Beginners","ML-For-Beginners 是由微软推出的一套系统化机器学习入门课程，旨在帮助零基础用户轻松掌握经典机器学习知识。这套课程将学习路径规划为 12 周，包含 26 节精炼课程和 52 道配套测验，内容涵盖从基础概念到实际应用的完整流程，有效解决了初学者面对庞大知识体系时无从下手、缺乏结构化指导的痛点。\n\n无论是希望转型的开发者、需要补充算法背景的研究人员，还是对人工智能充满好奇的普通爱好者，都能从中受益。课程不仅提供了清晰的理论讲解，还强调动手实践，让用户在循序渐进中建立扎实的技能基础。其独特的亮点在于强大的多语言支持，通过自动化机制提供了包括简体中文在内的 50 多种语言版本，极大地降低了全球不同背景用户的学习门槛。此外，项目采用开源协作模式，社区活跃且内容持续更新，确保学习者能获取前沿且准确的技术资讯。如果你正寻找一条清晰、友好且专业的机器学习入门之路，ML-For-Beginners 将是理想的起点。",84991,"2026-04-05T10:45:23",[14,51,52,53,15,54,26,13,55],"数据工具","视频","插件","其他","音频",{"id":57,"name":58,"github_repo":59,"description_zh":60,"stars":61,"difficulty_score":10,"last_commit_at":62,"category_tags":63,"status":16},3128,"ragflow","infiniflow\u002Fragflow","RAGFlow 是一款领先的开源检索增强生成（RAG）引擎，旨在为大语言模型构建更精准、可靠的上下文层。它巧妙地将前沿的 RAG 技术与智能体（Agent）能力相结合，不仅支持从各类文档中高效提取知识，还能让模型基于这些知识进行逻辑推理和任务执行。\n\n在大模型应用中，幻觉问题和知识滞后是常见痛点。RAGFlow 通过深度解析复杂文档结构（如表格、图表及混合排版），显著提升了信息检索的准确度，从而有效减少模型“胡编乱造”的现象，确保回答既有据可依又具备时效性。其内置的智能体机制更进一步，使系统不仅能回答问题，还能自主规划步骤解决复杂问题。\n\n这款工具特别适合开发者、企业技术团队以及 AI 研究人员使用。无论是希望快速搭建私有知识库问答系统，还是致力于探索大模型在垂直领域落地的创新者，都能从中受益。RAGFlow 提供了可视化的工作流编排界面和灵活的 API 接口，既降低了非算法背景用户的上手门槛，也满足了专业开发者对系统深度定制的需求。作为基于 Apache 2.0 协议开源的项目，它正成为连接通用大模型与行业专有知识之间的重要桥梁。",77062,"2026-04-04T04:44:48",[15,14,13,26,54],{"id":65,"github_repo":66,"name":67,"description_en":68,"description_zh":69,"ai_summary_zh":69,"readme_en":70,"readme_zh":71,"quickstart_zh":72,"use_case_zh":73,"hero_image_url":74,"owner_login":75,"owner_name":76,"owner_avatar_url":77,"owner_bio":78,"owner_company":79,"owner_location":80,"owner_email":79,"owner_twitter":79,"owner_website":79,"owner_url":81,"languages":79,"stars":82,"forks":83,"last_commit_at":84,"license":85,"difficulty_score":86,"env_os":87,"env_gpu":88,"env_ram":88,"env_deps":89,"category_tags":92,"github_topics":93,"view_count":23,"oss_zip_url":79,"oss_zip_packed_at":79,"status":16,"created_at":104,"updated_at":105,"faqs":106,"releases":135},3994,"Andrew-Jang\u002FRAGHub","RAGHub","A community-driven collection of RAG (Retrieval-Augmented Generation) frameworks, projects, and resources. Contribute and explore the evolving RAG ecosystem.","RAGHub 是一个由社区驱动的检索增强生成（RAG）工具与资源目录，旨在帮助开发者在快速演进的 AI 生态中轻松导航。面对每天层出不穷的新框架和项目，判断哪些工具真正实用、哪些只是短暂炒作变得愈发困难。RAGHub 通过系统性地收录和更新各类 RAG 框架、评估优化方案、数据预处理工具及实战项目，为用户提供一个清晰、前沿的参考指南。\n\n无论是刚接触 RAG 技术的新手，还是正在寻找最佳实践方案的资深工程师，都能在这里找到适合当前需求的技术选型建议。平台不仅列出了如 LangChain、Haystack 等主流框架，还持续追踪新兴项目，并附带活跃度指标与开源地址，方便用户快速评估其成熟度与社区支持情况。\n\nRAGHub 特别适合 AI 应用开发者、研究人员以及希望将大模型与企业数据结合的技术团队使用。其最大亮点在于“社区共建”模式——任何人都可提交新工具或更新现有条目，确保内容始终贴近实际开发场景。如果你正为如何选择 RAG 技术栈而困扰，不妨把 RAGHub 当作你的第一站，高效获取经过筛选的高质量资源，少走弯路，更快落地智能应用。","# RAGHub: A Directory of Tools for Retrieval-Augmented Generation (RAG)\n\nWelcome to **RAGHub**, a living collection of **new and emerging frameworks, projects, and resources** in the **Retrieval-Augmented Generation (RAG)** ecosystem. This is a **community-driven project for [r\u002FRAG](https:\u002F\u002Fwww.reddit.com\u002Fr\u002FRag\u002F)**, where we aim to catalog the rapid growth of RAG tools and projects that are pushing the boundaries of the field.\n\nEach day, it feels like a new tool or framework emerges, and choosing the right one is becoming more of an art than a science. Is the framework from three months ago still relevant? Or was it just hype, rehashing old concepts with a fresh look? **RAGHub exists to help you stay ahead of these changes**, providing a platform for the latest innovations in RAG.\n\n## How to Contribute\n\nThis is a community project, and **we welcome contributions from everyone**! If you’d like to add a new framework, project, or resource, please check out our [Contribution Guidelines](CONTRIBUTING.md) for details on how to get started.\n\n## Table of Contents\n\n- [RAGHub: A Directory of Tools for Retrieval-Augmented Generation (RAG)](#raghub-a-directory-of-tools-for-retrieval-augmented-generation-rag)\n - [How to Contribute](#how-to-contribute)\n - [Table of Contents](#table-of-contents)\n - [RAG Frameworks](#rag-frameworks)\n - [RAG Evaluation and Optimization Frameworks](#rag-evaluation-and-optimization-frameworks)\n - [RAG Engines](#rag-engines)\n - [RAG Data Preparation Frameworks](#rag-data-preparation-frameworks)\n - [RAG Projects](#rag-projects)\n - [RAG Resources and Sites](#rag-resources-and-sites)\n - [Model LeaderBoards](#model-leaderboards)\n - [License](#license)\n - [Join the Conversation](#join-the-conversation)\n\n## RAG Frameworks\n\n| Name | Description | Website | Github | Stars | Activity |\n| --------------------------- | ------------------------------------------------------------------ | ---------------------------------------------------------- | --------------------------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------- | ---------- |\n| Dcup Open-Source RAG-as-a-Service | Connect your app to user data in minutes with self-hostable RAG pipelines. | [Website](https:\u002F\u002Fdcup.dev) | [Github](https:\u002F\u002Fgithub.com\u002FDcup-dev\u002Fdcup) | [![GitHub](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FDcup-dev\u002Fdcup?color=5B5BD6)](https:\u002F\u002Fgithub.com\u002FDcup-dev\u002F) | 1h ago |\n| LangChain | Building applications with LLMs | [Website](https:\u002F\u002Flangchain.com) | [Github](https:\u002F\u002Fgithub.com\u002Flangchain-ai\u002Flangchain) | [![GitHub](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Flangchain-ai\u002Flangchain?color=5B5BD6)](https:\u002F\u002Fgithub.com\u002Flangchain-ai\u002Flangchain) | 9h ago |\n| Scout | Building apps with LLMs\u002Fvector databases\u002Fweb scraping | [Website](https:\u002F\u002Fscoutos.com) | [Github](https:\u002F\u002Fgithub.com\u002Fscoutos) | [![GitHub](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fscoutos?color=5B5BD6)](https:\u002F\u002Fgithub.com\u002Fscoutos) | 1h ago |\n| Haystack | A framework for building search engines using neural networks | [Website](https:\u002F\u002Fhaystack.deepset.ai) | [Github](https:\u002F\u002Fgithub.com\u002Fdeepset-ai\u002Fhaystack) | [![GitHub](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fdeepset-ai\u002Fhaystack?color=5B5BD6)](https:\u002F\u002Fgithub.com\u002Fdeepset-ai\u002Fhaystack) | Last week |\n| LlamaIndex | A framework for building data-driven LLM applications | [Website](https:\u002F\u002Fwww.llamaindex.ai\u002F) | [Github](https:\u002F\u002Fgithub.com\u002Frun-llama\u002Fllama_index) | [![GitHub](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Frun-llama\u002Fllama_index?color=5B5BD6)](https:\u002F\u002Fgithub.com\u002Frun-llama\u002Fllama_index) | 7h ago |\n| BentoML | Build Inference APIs, LLM apps, Multi-model chains, RAG | [Website](https:\u002F\u002Fwww.bentoml.com\u002F) | [Github](https:\u002F\u002Fgithub.com\u002Fbentoml\u002FBentoML) | [![GitHub](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fbentoml\u002FBentoML?color=5B5BD6)](https:\u002F\u002Fgithub.com\u002Fbentoml\u002FBentoML) | 1h ago |\n| Contextual AI | End-to-end RAG including document understanding, retrieval, grounded generation, and evaluation | [Website](https:\u002F\u002Fcontextual.ai) | [GitHub](https:\u002F\u002Fgithub.com\u002FContextualAI\u002Fexamples) | ![GitHub stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FContextualAI\u002Fexamples?style=social) | -- |\n| LightRAG | Simple and fast Retrieval-Augmented Generation | [Website](https:\u002F\u002Farxiv.org\u002Fabs\u002F2410.05779) | [Github](https:\u002F\u002Fgithub.com\u002FHKUDS\u002FLightRAG) | [![GitHub](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FHKUDS\u002FLightRAG?color=5B5BD6)](https:\u002F\u002Fgithub.com\u002FHKUDS\u002FLightRAG) | 1d ago |\n| Swarm by OpenAI | Educational framework for lightweight multi-agent orchestration | - | [Github](https:\u002F\u002Fgithub.com\u002Fopenai\u002Fswarm) | [![GitHub](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fopenai\u002Fswarm?color=5B5BD6)](https:\u002F\u002Fgithub.com\u002Fopenai\u002Fswarm) | 1d ago |\n| Langroid | Python framework to easily build LLM-powered applications | [Website](https:\u002F\u002Flangroid.github.io\u002Flangroid\u002F) | [Github](https:\u002F\u002Fgithub.com\u002Flangroid\u002Flangroid) | [![GitHub](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Flangroid\u002Flangroid?color=5B5BD6)](https:\u002F\u002Fgithub.com\u002Flangroid\u002Flangroid) | 10h ago |\n| NeMo-Guardrails | Add programmable guardrails to LLM-based applications | [Website](https:\u002F\u002Fdocs.nvidia.com\u002Fnemo-guardrails\u002F) | [Github](https:\u002F\u002Fgithub.com\u002FNVIDIA\u002FNeMo-Guardrails) | [![GitHub](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FNVIDIA\u002FNeMo-Guardrails?color=5B5BD6)](https:\u002F\u002Fgithub.com\u002FNVIDIA\u002FNeMo-Guardrails) | Last week |\n| Swiftide | A Rust library for building fast, streaming applications with LLMs | [Website](https:\u002F\u002Fswiftide.rs) | [GitHub](https:\u002F\u002Fgithub.com\u002Fbosun-ai\u002Fswiftide) | ![GitHub stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fbosun-ai\u002Fswiftide?style=social) | 1h ago |\n| Korvus | The entire RAG pipeline in a single database query | [Website](https:\u002F\u002Fpostgresml.org) | [GitHub](https:\u002F\u002Fgithub.com\u002Fpostgresml\u002Fkorvus) | ![GitHub stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fpostgresml\u002Fkorvus?style=social) | Last month |\n| semantic-router | A framework for routing LLM requests using semantic vectors | [Website](https:\u002F\u002Fwww.aurelio.ai\u002Fsemantic-router) | [GitHub](https:\u002F\u002Fgithub.com\u002Faurelio-labs\u002Fsemantic-router) | ![GitHub stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Faurelio-labs\u002Fsemantic-router?style=social) | 4h ago |\n| AWS Bedrock Knowledge Bases | Service to build, scale, and deploy RAG-powered applications | [Website](https:\u002F\u002Faws.amazon.com\u002Fbedrock\u002Fknowledge-bases\u002F) | - | - | 1h ago |\n| langflow | Build, scale, and deploy RAG and multi-agent AI apps | [Website](https:\u002F\u002Fwww.langflow.org\u002F) | [GitHub](https:\u002F\u002Fgithub.com\u002Flangflow-ai\u002Flangflow) | ![GitHub stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Flangflow-ai\u002Flangflow?style=social) | 1h ago |\n| dspy | Build language model apps with modular programming | [Website](https:\u002F\u002Fdspy-docs.vercel.app\u002F) | [GitHub](https:\u002F\u002Fgithub.com\u002Fstanfordnlp\u002Fdspy) | ![GitHub stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fstanfordnlp\u002Fdspy?style=social) | 13h ago |\n| mem0 | The Memory layer for your AI apps | [Website](https:\u002F\u002Fmem0.ai\u002F) | [GitHub](https:\u002F\u002Fgithub.com\u002Fmem0ai\u002Fmem0) | ![GitHub stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fmem0ai\u002Fmem0?style=social) | 2h ago |\n| RAGLite | A Python package for building RAG applications | [Website](https:\u002F\u002Fsuperlinear.eu) | [GitHub](https:\u002F\u002Fgithub.com\u002Fsuperlinear-ai\u002Fraglite) | ![GitHub stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fsuperlinear-ai\u002Fraglite?style=social) | 18h ago |\n| cognee | Memory framework for building GraphRAG applications | [Website](https:\u002F\u002Fwww.cognee.ai) | [GitHub](https:\u002F\u002Fgithub.com\u002Ftopoteretes\u002Fcognee) | ![GitHub stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Ftopoteretes\u002Fcognee?style=social) | 2h ago |\n| ragbits | Building blocks for rapid development of GenAI applications | [Website](https:\u002F\u002Fragbits.deepsense.ai) | [GitHub](https:\u002F\u002Fgithub.com\u002Fdeepsense-ai\u002Fragbits) | ![GitHub stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fdeepsense-ai\u002Fragbits?style=social) | 1h ago |\n| Interchange | End-to-end API for RAG, from document upload to search | [Website](https:\u002F\u002Fwww.getinterchange.io) | [GitHub](https:\u002F\u002Fgithub.com\u002Fgetinterchange) | [![GitHub](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fgetinterchange?color=5B5BD6)](https:\u002F\u002Fgithub.com\u002Fgetinterchange) | 1h ago |\n| ZeroEntropy | Rerankers, embeddings and end-to-end retrieval API | [Website](https:\u002F\u002Fdocs.zeroentropy.dev) | [GitHub](https:\u002F\u002Fgithub.com\u002Forgs\u002Fzeroentropy-ai\u002Frepositories) | - | 1h ago |\n| memori | Multi-Agent Memory Engine that gives your AI agents human-like memory | [Website](https:\u002F\u002Fmemori.gibsonai.com\u002F) | [GitHub](https:\u002F\u002Fgithub.com\u002Fgibsonai\u002Fmemori) | ![GitHub stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fgibsonai\u002Fmemori?style=social) | 3d ago |\n\n\n## RAG Evaluation and Optimization Frameworks\n\n| Name | Description | Website | GitHub | Stars | Activity |\n| ------------ | ------------------------------------------------------------------------------------------------------ | ---------------------------------------------------------------------------- | ------------------------------------------------------ | ------------------------------------------------------------------------------------------- | -------- |\n| Trulens | Measures and enhance LLM app quality with feedback functions for scalable evaluation | [Website](https:\u002F\u002Fwww.trulens.org\u002F) | [GitHub](https:\u002F\u002Fgithub.com\u002Ftruera\u002Ftrulens) | ![GitHub stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Ftruera\u002Ftrulens?style=social) | 11h ago |\n| Phoenix | AI observability platform designed for experimentation, evaluation, and troubleshooting | [Website](https:\u002F\u002Fgithub.com\u002FArize-ai\u002Fphoenix) | [GitHub](https:\u002F\u002Fgithub.com\u002FArize-ai\u002Fphoenix) | ![GitHub stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FArize-ai\u002Fphoenix?style=social) | 1d ago |\n| ragas | Evaluates and quantifies the performance of RAG pipelines that enhance LLM context with external data | [Website](https:\u002F\u002Fdocs.ragas.io\u002Fen\u002Fstable\u002F) | [GitHub](https:\u002F\u002Fgithub.com\u002Fexplodinggradients\u002Fragas) | ![GitHub stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fexplodinggradients\u002Fragas?style=social) | 3h ago |\n| LMUnit | Language model optimized for evaluating natural language unit tests | [Website](https:\u002F\u002Fdocs.contextual.ai\u002Fapi-reference\u002Flmunit\u002Flmunit) | - | - | -- |\n| Deepchecks | Continuous validation of AI & ML models, detecting data drift and model issues | [Website](https:\u002F\u002Fdocs.deepchecks.com\u002Fstable) | [GitHub](https:\u002F\u002Fgithub.com\u002Fdeepchecks\u002Fdeepchecks) | ![GitHub stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fdeepchecks\u002Fdeepchecks?style=social) | 8m ago |\n| AutoRAG | End-to-end RAG optimization: parsing, chunking, evaluation dataset creation, and pipeline deployment | [Website](https:\u002F\u002Fauto-rag.com\u002F) | [GitHub](https:\u002F\u002Fgithub.com\u002FMarker-Inc-Korea\u002FAutoRAG) | ![GitHub stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FMarker-Inc-Korea\u002FAutoRAG?style=social) | 1h ago |\n| evalmy.ai | Fine-tuned lightweight RAG evaluation service + Python client library | [Website](https:\u002F\u002Fwww.evalmy.ai\u002F) | [GitHub](https:\u002F\u002Fgithub.com\u002Fevalmy-ai\u002Fevalmyai-python) | ![GitHub stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fevalmy-ai\u002Fevalmyai-python?style=social) | -- |\n| TextGrad | A framework for LLM-based text optimization, focusing on reducing hallucinations and improving prompts | [Website](https:\u002F\u002Ftextgrad.com\u002F) | [GitHub](https:\u002F\u002Fgithub.com\u002Fzou-group\u002Ftextgrad) | ![GitHub stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fzou-group\u002Ftextgrad?style=social) | 24h ago |\n| langfuse | Traces, evals, prompt management, and metrics to debug and improve your LLM application. | [Website](https:\u002F\u002Flangfuse.com\u002F) | [GitHub](https:\u002F\u002Fgithub.com\u002Flangfuse\u002Flangfuse) | ![GitHub stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Flangfuse\u002Flangfuse?style=social) | 1h ago |\n| Vectara HHEM | Hallucination evaluation model for RAG | [Huggingface](https:\u002F\u002Fhuggingface.co\u002Fvectara\u002Fhallucination_evaluation_model) | -- | -- | -- |\n| StepsTrack | An Observability tool built to track, inspect, and visualize every steps in a pipeline | - | [GitHub](https:\u002F\u002Fgithub.com\u002Flokwkin\u002Fsteps-track) | ![GitHub stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Flokwkin\u002Fsteps-track?style=social) | 15h ago |\n| syftr | Multi-objective end-to-end agentic RAG optimization. | - | [GitHub](https:\u002F\u002Fgithub.com\u002Fdatarobot\u002Fsyftr) | ![GitHub stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fdatarobot\u002Fsyftr?style=social) | 1h ago |\n| zbench | Annotation and evaluation framework for retrieval and reranking | [Website](https:\u002F\u002Fdocs.zeroentropy.dev) | [GitHub](https:\u002F\u002Fgithub.com\u002FZeroEntropy-AI\u002Fzbench) | - | 1h ago |\n| rag-select | End-to-end RAG architecture evaluation\u002Foptimization across RAG tool offerings. | [Website](https:\u002F\u002Fuseconclude.com\u002Fengineering\u002Frag-select) | [GitHub](https:\u002F\u002Fgithub.com\u002Fconclude-ai\u002Frag-select) | ![GitHub stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fconclude-ai\u002Frag-select?style=social) | 1h ago |\n\n## RAG Engines\n\n| Name | Description | Website | GitHub | Stars | Activity |\n| -------------------------- | ---------------------------------------------------------------------------------------------- | ----------------------------------------------------------------------------- | ------------------------------------------------------------- | ------------------------------------------------------------------------------------------------ | ------------ |\n| Agentset | Open-source agentic RAG platform. | [Website](https:\u002F\u002Fagentset.ai) | [GitHub](https:\u002F\u002Fgithub.com\u002Fagentset-ai\u002Fagentset) | ![GitHub stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fagentset-ai\u002Fagentset?style=social) | 1d ago |\n| Engramic | RAG engine focused on long-term memory and advanced context management | [Website](https:\u002F\u002Fengramic.org) | [Github](https:\u002F\u002Fgithub.com\u002Fengramic\u002Fengramic) | ![GitHub stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fengramic\u002Fengramic?style=social) | 2h ago |\n| TrustGraph | LLM Agnostic Agent Development Platform | [Website](https:\u002F\u002Ftrustgraph.ai) | [GitHub](https:\u002F\u002Fgithub.com\u002Ftrustgraph-ai\u002Ftrustgraph) | ![GitHub stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Ftrustgraph-ai\u002Ftrustgraph?style=social) | 2d ago |\n| R2R | The Elasticsearch for RAG, helps you quickly build and launch scalable RAG solutions | [Website](https:\u002F\u002Fr2r-docs.sciphi.ai\u002Fintroduction) | [GitHub](https:\u002F\u002Fgithub.com\u002FSciPhi-AI\u002FR2R) | ![GitHub stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FSciPhi-AI\u002FR2R?style=social) | 6h ago |\n| RAGFlow | Open-source RAG engine based on deep document understanding | [Website](https:\u002F\u002Fragflow.io) | [GitHub](https:\u002F\u002Fgithub.com\u002Finfiniflow\u002Fragflow) | ![GitHub stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Finfiniflow\u002Fragflow?style=social) | 1h ago |\n| Liquid Index | The Unified RAG Platform. One API. Every Tool You Need | [Website](https:\u002F\u002Fliquidindex.dev) | - | - | 1h ago |\n| Vertex AI Knowledge Engine | A data framework for context-augmented LLM applications | [Website](https:\u002F\u002Fcloud.google.com\u002Fvertex-ai\u002Fgenerative-ai\u002Fdocs\u002Frag-overview) | - | - | 1d ago |\n| Embedchain | Open Source Framework for personalizing LLM responses under 10 lines of code | [Website](https:\u002F\u002Fdocs.embedchain.ai\u002Fget-started\u002Fquickstart) | [GitHub](https:\u002F\u002Fgithub.com\u002Fmem0ai\u002Fmem0\u002Ftree\u002Fmain\u002Fembedchain) | ![GitHub stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fmem0ai\u002Fmem0?style=social) | Last week |\n| txtai | All-in-one embeddings database for semantic search, LLM orchestration, and RAG workflows | [Website](https:\u002F\u002Fneuml.github.io\u002Ftxtai\u002F) | [GitHub](https:\u002F\u002Fgithub.com\u002Fneuml\u002Ftxtai) | ![GitHub stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fneuml\u002Ftxtai?style=social) | Last week |\n| dsRAG | High-performance retrieval engine for unstructured data | - | [GitHub](https:\u002F\u002Fgithub.com\u002FD-Star-AI\u002FdsRAG) | ![GitHub stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FD-Star-AI\u002FdsRAG?style=social) | Last week |\n| Flash-Rank | Use Pairwise or Listwise rerankers to improve search accuracy before passing to LLMs. | - | [GitHub](https:\u002F\u002Fgithub.com\u002FPrithivirajDamodaran\u002FFlashRank) | ![GitHub stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FPrithivirajDamodaran\u002FFlashRank?style=social) | 2w ago |\n| Graphlit | API-first platform for building knowledge-driven AI applications and agents | [Website](https:\u002F\u002Fwww.graphlit.com) | [GitHub](https:\u002F\u002Fgithub.com\u002Fgraphlit) | ![GitHub stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fgraphlit?style=social) | 8h ago |\n| rag-citation | Combines RAG with automatic citation generation to enhance content credibility | [Website](https:\u002F\u002Fpypi.org\u002Fproject\u002Frag-citation\u002F) | [GitHub](https:\u002F\u002Fgithub.com\u002Frahulanand1103\u002Frag-citation) | ![GitHub stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Frahulanand1103\u002Frag-citation?style=social) | Last week |\n| PostgresML | Postgres + GPUs with functions for chunking, embedding, transforming and ranking | [Website](https:\u002F\u002Fpostgresml.org) | [GitHub](https:\u002F\u002Fgithub.com\u002Fpostgresml\u002Fpostgresml) | ![GitHub stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fpostgresml\u002Fpostgresml?style=social) | Yesterday |\n| chainlit | Build production-ready Conversational AI applications in minutes, not weeks | [Website](https:\u002F\u002Fchainlit.io\u002F) | [GitHub](https:\u002F\u002Fgithub.com\u002FChainlit\u002Fchainlit) | ![GitHub stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FChainlit\u002Fchainlit?style=social) | 24h ago |\n| pathway | Python ETL framework for stream processing, real-time analytics, LLM pipelines, and RAG. | [Website](https:\u002F\u002Fpathway.com\u002F) | [GitHub](https:\u002F\u002Fgithub.com\u002Fpathwaycom\u002Fpathway) | ![GitHub stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fpathwaycom\u002Fpathway?style=social) | 7h ago |\n| cognita | RAG framework for modular, open-source production apps. | [Website](https:\u002F\u002Fcognita.truefoundry.com\u002F) | [GitHub](https:\u002F\u002Fgithub.com\u002Ftruefoundry\u002Fcognita) | ![GitHub stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Ftruefoundry\u002Fcognita?style=social) | 2 days ago |\n| FlashRAG | A Python Toolkit for Efficient RAG Research | - | [GitHub](https:\u002F\u002Fgithub.com\u002FRUC-NLPIR\u002FFlashRAG) | ![GitHub stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FRUC-NLPIR\u002FFlashRAG?style=social) | 3h ago |\n| RAGatouille | Easily train and use advanced retrieval methods in any RAG pipeline. | - | [GitHub](https:\u002F\u002Fgithub.com\u002FAnswerDotAI\u002FRAGatouille) | ![GitHub stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FAnswerDotAI\u002FRAGatouille?style=social) | 4 months ago |\n| pgai | A suite of tools to develop RAG, semantic search, and other AI applications just in PostgreSQL | [Website](https:\u002F\u002Fwww.timescale.com\u002Fai) | [GitHub](https:\u002F\u002Fgithub.com\u002Ftimescale\u002Fpgai) | ![GitHub stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Ftimescale\u002Fpgai?style=social) | 10h ago |\n| Vectara | The trusted RAG platform for quickly building AI assistants and agents. | [Website](https:\u002F\u002Fwww.vectara.com\u002F) | [GitHub](https:\u002F\u002Fgithub.com\u002Fvectara\u002F) | - | - |\n| mode | RAG framework with expert models, smart clustering,and efficient retrieval for small datasets. | - | [GitHub](https:\u002F\u002Fgithub.com\u002Frahulanand1103\u002Fmode) |![GitHub stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Frahulanand1103\u002Fmode?style=social) | 2 days ago |\n| haiku.rag | Open-Source RAG framework with monitoring, CLI, search, Q\u002FA, MCP support on SQLite. | - | [Github](https:\u002F\u002Fgithub.com\u002Fggozad\u002Fhaiku.rag) |![GitHub stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fggozad\u002Fhaiku.rag?style=social) | 3h ago |\n| ZeroEntropy AI | Open-Weight Rerankers, Embeddings and End-to-End Retrieval API | [Website](https:\u002F\u002Fdocs.zeroentropy.dev) | [Github](https:\u002F\u002Fgithub.com\u002FZeroEntropy-AI) | - | 3h ago |\n\n## RAG Data Preparation Frameworks\n\n| Name | Description | Website | GitHub | Stars | Activity |\n| --------- | ---------------------------------- | -------------------------------- | --------------------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------- | -------- |\n| CocoIndex | ETL framework to build fresh index | [Website](https:\u002F\u002Fcocoindex.io\u002F) | [Github](https:\u002F\u002Fgithub.com\u002Fcocoindex-io\u002Fcocoindex) | [![GitHub](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fcocoindex-io\u002Fcocoindex?color=5B5BD6)](https:\u002F\u002Fgithub.com\u002Fcocoindex-io\u002Fcocoindex) | 1h ago |\n| Gitana.io | Content platform for editorial approval and scheduled deployment of trained data sets to RAG vector DBs | [Website](https:\u002F\u002Fgitana.io\u002F) | - | - | - |\n| Chonkie | No-nonsense, lightweight and fast RAG chunking library | [Website](https:\u002F\u002Fchonkie.ai\u002F) | [GitHub](https:\u002F\u002Fgithub.com\u002Fchonkie-inc\u002Fchonkie) | [![GitHub stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fchonkie-inc\u002Fchonkie?style=social)](https:\u002F\u002Fgithub.com\u002Fchonkie-inc\u002Fchonkie) | 1h ago |\n\n## RAG Projects\n\n| Name | Description | Website | GitHub | Stars | Activity |\n|------------------------------------------|--------------------------------------------------------------------------------------------------------------------| ----------------------------------------------------------------------------------------- | ----------------------------------- | ------------------------- | --------- |\n| LlamaParse | GenAI-native document parsing platform | [Website](https:\u002F\u002Fdocs.llamaindex.ai\u002Fen\u002Fstable\u002Fllama_cloud\u002Fllama_parse\u002F) | [GitHub](https:\u002F\u002Fgithub.com\u002Frun-llama\u002Fllama_parse) | ![GitHub stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Frun-llama\u002Fllama_parse?style=social) | 2d ago |\n| Langchain-extract | Web server to extract information from text and files using LLMs | [Website](https:\u002F\u002Fpython.langchain.com\u002Fv0.1\u002Fdocs\u002Fuse_cases\u002Fextraction\u002F) | [GitHub](https:\u002F\u002Fgithub.com\u002Flangchain-ai\u002Flangchain-extract) | ![GitHub stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Flangchain-ai\u002Flangchain-extract?style=social) | 4m ago |\n| Needle | Production-ready RAG pipelines out of the box. | [Website](https:\u002F\u002Fneedle-ai.com\u002F) | [GitHub](https:\u002F\u002Fgithub.com\u002Foeken\u002Fneedle-python) | ![GitHub stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Foeken\u002Fneedle-python?style=social) | 1h ago |\n| Unstructured.io | Build custom preprocessing pipelines for labeling, training, or production ML | [Website](https:\u002F\u002Funstructured.io\u002F) | [GitHub](https:\u002F\u002Fgithub.com\u002FUnstructured-IO\u002Funstructured) | ![GitHub stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FUnstructured-IO\u002Funstructured?style=social) | 3d ago |\n| Verba | RAG chatbot powered by Weaviate | [Website](https:\u002F\u002Fverba.weaviate.io\u002F) | [GitHub](https:\u002F\u002Fgithub.com\u002Fweaviate\u002FVerba) | ![GitHub stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fweaviate\u002FVerba?style=social) | 2w ago |\n| Unstract | No-code platform to launch APIs and ETL Pipelines to structure unstructured documents | [Website](https:\u002F\u002Funstract.com\u002F) | [GitHub](https:\u002F\u002Fgithub.com\u002FZipstack\u002Funstract) | ![GitHub stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FZipstack\u002Funstract?style=social) | 4h ago |\n| Humata.ai | Ask questions across all of your document files | [Website](https:\u002F\u002Fwww.humata.ai\u002F) | - | - | 4h ago |\n| Kiln | Build AI systems with evals, RAG, agents, fine-tuning, and synthetic data | [Website](https:\u002F\u002Fkiln.tech) | [GitHub](https:\u002F\u002Fgithub.com\u002FKiln-AI\u002FKiln) | ![GitHub stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FKiln-AI\u002FKiln?style=social) | 12h ago |\n| Ragie.ai | Fully managed RAG-as-a-Service for developers. | [Website](https:\u002F\u002Fwww.ragie.ai\u002F) | [GitHub](https:\u002F\u002Fgithub.com\u002Fragieai) | - | 12h ago |\n| Reducto | Parses complex documents and creates LLM-ready inputs | [Website](https:\u002F\u002Freducto.ai\u002F) | [GitHub](https:\u002F\u002Fgithub.com\u002Freductoai) | - | 2w ago |\n| Midship | Extract document data straight into your spreadsheet\u002FERP\u002FCRM | [Website](https:\u002F\u002Fmidship.ai\u002F) | - | - | - |\n| DocuPanda | Convert documents into a structured, standard set of fields and values | [Website](https:\u002F\u002Fwww.docupanda.io\u002F) | - | - | - |\n| contextual-doc-retrieval-opneai-reranker | Using GPT-4 and Cohere for query expansion and re-ranking with BM25 | - | [GitHub](https:\u002F\u002Fgithub.com\u002Flesteroliver911\u002Fcontextual-doc-retrieval-opneai-reranker) | ![GitHub stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Flesteroliver911\u002Fcontextual-doc-retrieval-opneai-reranker?style=social) | Last week |\n| Raggenie | Low-code platform to build custom RAG-based AI applications | [Website](https:\u002F\u002Fwww.raggenie.com) | [GitHub](https:\u002F\u002Fgithub.com\u002Fsirocco-ventures\u002Fraggenie) | ![GitHub stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fsirocco-ventures\u002Fraggenie?style=social) | 10h ago |\n| Chunkr | Vision model-based PDF chunking and OCR, optimized for fast processing of large datasets | [Website](https:\u002F\u002Fchunkr.ai) | [GitHub](https:\u002F\u002Fgithub.com\u002Flumina-ai-inc\u002Fchunkr) | ![GitHub stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Flumina-ai-inc\u002Fchunkr?style=social) | 11h ago |\n| tldw | Open-source project similar to NotebookLM | [Website](https:\u002F\u002Ftldwproject.com) | [GitHub](https:\u002F\u002Fgithub.com\u002Frmusser01\u002Ftldw) | ![GitHub stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Frmusser01\u002Ftldw?style=social) | Yesterday |\n| Cerbos | Access control for RAG and LLMs. | [Website](https:\u002F\u002Fsolutions.cerbos.dev\u002Fauthorization-in-rag-based-ai-systems-with-cerbos) | [GitHub](https:\u002F\u002Fgithub.com\u002Fcerbos\u002Fcerbos) | ![GitHub stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fcerbos\u002Fcerbos?style=social) | 14h ago |\n| extractous | Extremely fast data extraction for your AI applications | [Website](https:\u002F\u002Fwww.extractous.com\u002F) | [GitHub](https:\u002F\u002Fgithub.com\u002Fyobix-ai\u002Fextractous) | ![GitHub stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fyobix-ai\u002Fextractous?style=social) | - |\n| SWIRL | AI search & RAG for your workplace. Get AI insights from your company's knowledge instantly. | [Website](https:\u002F\u002Fwww.swirlaiconnect.com\u002F) | [GitHub](https:\u002F\u002Fgithub.com\u002Fswirlai\u002Fswirl-search) | ![GitHub stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fswirlai\u002Fswirl-search?style=social) | 2w ago |\n| ChatDOC PDF Parser | Precision PDF parsing that transforms documents into flawless structured data for RAG systems. | [Website](https:\u002F\u002Fpdfparser.io\u002F?src=github) | - | - | - |\n| Gurubase | Create AI-powered Q&A assistants by indexing websites, PDF documents, YouTube videos, and GitHub code repositories. | [Website](https:\u002F\u002Fgurubase.io) | [GitHub](https:\u002F\u002Fgithub.com\u002FGurubase\u002Fgurubase) | ![GitHub stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FGurubase\u002Fgurubase?style=social) | 1d ago |\n| Archive Agent | Open-source semantic file tracker with OCR + AI search. Smart indexer with RAG engine. | - | [GitHub](https:\u002F\u002Fgithub.com\u002FshredEngineer\u002FArchive-Agent) | ![GitHub stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FshredEngineer\u002FArchive-Agent?style=social) | - |\n| MidrasAI | Simple API for Colpali, a multi-modal retrieval model. | - | [Github](https:\u002F\u002Fgithub.com\u002Fajac-zero\u002Fmidrasai) | ![GitHub stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fajac-zero\u002Fmidrasai?style=social) | 6m ago |\n| EmbeddingBridge | Version control and migration tool for embeddings | - | [Github](https:\u002F\u002Fgithub.com\u002FProgramComputer\u002FEmbeddingBridge) | ![Github stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FProgramComputer\u002FEmbeddingBridge?style=social)|-|\n| Stream-Rag-Agent | Streaming RAG Agent for Kafka | - | [Github](https:\u002F\u002Fgithub.com\u002Fonurbaran\u002Fstream-rag-agent) | |-|\n| zchunk | Open-Source efficient LLM-based chunking | [Website](https:\u002F\u002Fdocs.zeroentropy.dev) | [Github](https:\u002F\u002Fgithub.com\u002FZeroEntropy-AI\u002Fzchunk) | - | 2h ago\n| hydrot | A production-ready Retrieval-Augmented Generation (RAG) system designed for enterprise documentation, with first-class support for markdown content. Built with a microservices architecture for horizontal scaling and flexibility. | - | [Github](https:\u002F\u002Fgithub.com\u002Frjxby\u002Fhydrot) | ![Github stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Frjxby\u002Fhydrot?style=social)|-|\n| Tensorlake | Document parsing for citable, traceable RAG systems. Extracts text, tables, and images while maintaining complex layouts, attaching bounding box coordinates, and supporting structured output | [Website](https:\u002F\u002Fwww.tensorlake.ai\u002F) | [Github](https:\u002F\u002Fgithub.com\u002Ftensorlakeai) | ![Github stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Ftensorlakeai\u002Findexify?style=social) | 2d ago|\n\n## RAG Resources and Sites\n\n| Site\u002FArticle | Description | Link |\n| -------------------- | ----------------------------------------------------------------------------- | -------------------------------------------------------------- |\n| Contextual Retrieval | Anthropic introducing Contextual Retrieval | [Website](https:\u002F\u002Fwww.anthropic.com\u002Fnews\u002Fcontextual-retrieval) |\n| Open-RAG | Enhanced Retrieval-Augmented Reasoning with Open-Source Large Language Models | [Website](https:\u002F\u002Farxiv.org\u002Fabs\u002F2410.01782) |\n| ColPali | Efficient Document Retrieval with Vision Language Models | [Website](https:\u002F\u002Farxiv.org\u002Fabs\u002F2407.01449) |\n| RAG_Techniques | Showcases various advanced techniques for RAG systems | [Website](https:\u002F\u002Fgithub.com\u002FNirDiamant\u002FRAG_Techniques) |\n| GenAI_Agents | Tutorials and implementations for various AI Agent techniques | [Website](https:\u002F\u002Fgithub.com\u002FNirDiamant\u002FGenAI_Agents) |\n| RAG From Scratch | Official LangChain guide for building a RAG pipeline from scratch | [Website](https:\u002F\u002Fgithub.com\u002Flangchain-ai\u002Frag-from-scratch) |\n\n\n## Model LeaderBoards\n\n| Name | Description | Link |\n| --------------------------------- | ---------------------------------- | --------------------------------------------------------------- |\n| Artificial Analysis | LLM Comparison | [Website](https:\u002F\u002Fartificialanalysis.ai) |\n| HuggingFace\u002Fmteb | Embedding models leaderboard | [Website](https:\u002F\u002Fhuggingface.co\u002Fspaces\u002Fmteb\u002Fleaderboard) |\n| Vectara Hallucination Leaderboard | Hallucination leaderboard for LLMs | [Website](https:\u002F\u002Fgithub.com\u002Fvectara\u002Fhallucination-leaderboard) |\n\n> If you're looking for mainstream RAG frameworks and techniques, check out the excellent repository by Nir Diamant: [RAG Techniques](https:\u002F\u002Fgithub.com\u002FNirDiamant\u002FRAG_Techniques). This repository focuses on more established tools and methods that have already gained traction in the community.\n\n## License\n\nThis project is licensed under the **MIT License**. See the [LICENSE](LICENSE) file for details.\n\n## Join the Conversation\n\nThis project is part of the [r\u002FRAG](https:\u002F\u002Fwww.reddit.com\u002Fr\u002FRag\u002F) community. Have feedback or suggestions? Feel free to open an issue, start a discussion, or join the conversation on our [Discord server](https:\u002F\u002Fdiscord.gg\u002Fnn92wC5QmN)! We want to make this repository a valuable resource for everyone exploring the RAG ecosystem, and your input is crucial.\n","# RAGHub：检索增强生成（RAG）工具目录\n\n欢迎来到**RAGHub**，这是一个不断更新的**检索增强生成（RAG）生态系统中新兴框架、项目和资源**的集合。这是一个由社区驱动的项目，面向[r\u002FRAG](https:\u002F\u002Fwww.reddit.com\u002Fr\u002FRag\u002F)社区，旨在梳理快速发展的RAG工具与项目，推动该领域的边界不断拓展。\n\n每天似乎都有新的工具或框架涌现，而如何选择合适的工具正逐渐从一门科学演变为一种艺术。三个月前的那个框架现在还适用吗？还是仅仅是一阵炒作，用新包装重新演绎旧概念？**RAGHub的存在就是为了帮助您紧跟这些变化**，为您提供RAG领域最新创新成果的展示平台。\n\n## 如何贡献\n\n这是一个社区共建的项目，**我们欢迎所有人的参与和贡献**！如果您希望添加新的框架、项目或资源，请查看我们的[贡献指南](CONTRIBUTING.md)，了解具体的操作步骤。\n\n## 目录\n\n- [RAGHub：检索增强生成（RAG）工具目录](#raghub-a-directory-of-tools-for-retrieval-augmented-generation-rag)\n - [如何贡献](#how-to-contribute)\n - [目录](#table-of-contents)\n - [RAG框架](#rag-frameworks)\n - [RAG评估与优化框架](#rag-evaluation-and-optimization-frameworks)\n - [RAG引擎](#rag-engines)\n - [RAG数据准备框架](#rag-data-preparation-frameworks)\n - [RAG项目](#rag-projects)\n - [RAG资源与网站](#rag-resources-and-sites)\n - [模型排行榜](#model-leaderboards)\n - [许可证](#license)\n - [加入讨论](#join-the-conversation)\n\n## RAG框架\n\n| 名称 | 描述 | 官网 | Github | 星数 | 活跃度 |\n| --------------------------- | ------------------------------------------------------------------ | ---------------------------------------------------------- | --------------------------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------- | ---------- |\n| Dcup 开源 RAG 即服务 | 通过可自托管的 RAG 流程，可在几分钟内将您的应用连接到用户数据。 | [官网](https:\u002F\u002Fdcup.dev) | [Github](https:\u002F\u002Fgithub.com\u002FDcup-dev\u002Fdcup) | [![GitHub](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FDcup-dev\u002Fdcup?color=5B5BD6)](https:\u002F\u002Fgithub.com\u002FDcup-dev\u002F) | 1小时前 |\n| LangChain | 使用大语言模型构建应用程序 | [官网](https:\u002F\u002Flangchain.com) | [Github](https:\u002F\u002Fgithub.com\u002Flangchain-ai\u002Flangchain) | [![GitHub](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Flangchain-ai\u002Flangchain?color=5B5BD6)](https:\u002F\u002Fgithub.com\u002Flangchain-ai\u002Flangchain) | 9小时前 |\n| Scout | 使用大语言模型\u002F向量数据库\u002F网页抓取构建应用 | [官网](https:\u002F\u002Fscoutos.com) | [Github](https:\u002F\u002Fgithub.com\u002Fscoutos) | [![GitHub](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fscoutos?color=5B5BD6)](https:\u002F\u002Fgithub.com\u002Fscoutos) | 1小时前 |\n| Haystack | 用于构建基于神经网络搜索引擎的框架 | [官网](https:\u002F\u002Fhaystack.deepset.ai) | [Github](https:\u002F\u002Fgithub.com\u002Fdeepset-ai\u002Fhaystack) | [![GitHub](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fdeepset-ai\u002Fhaystack?color=5B5BD6)](https:\u002F\u002Fgithub.com\u002Fdeepset-ai\u002Fhaystack) | 上周 |\n| LlamaIndex | 构建数据驱动的大语言模型应用的框架 | [官网](https:\u002F\u002Fwww.llamaindex.ai\u002F) | [Github](https:\u002F\u002Fgithub.com\u002Frun-llama\u002Fllama_index) | [![GitHub](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Frun-llama\u002Fllama_index?color=5B5BD6)](https:\u002F\u002Fgithub.com\u002Frun-llama\u002Fllama_index) | 7小时前 |\n| BentoML | 构建推理 API、大语言模型应用、多模型链和 RAG | [官网](https:\u002F\u002Fwww.bentoml.com\u002F) | [Github](https:\u002F\u002Fgithub.com\u002Fbentoml\u002FBentoML) | [![GitHub](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fbentoml\u002FBentoML?color=5B5BD6)](https:\u002F\u002Fgithub.com\u002Fbentoml\u002FBentoML) | 1小时前 |\n| Contextual AI | 端到端 RAG，包括文档理解、检索、 grounded 生成和评估 | [官网](https:\u002F\u002Fcontextual.ai) | [GitHub](https:\u002F\u002Fgithub.com\u002FContextualAI\u002Fexamples) | ![GitHub 星数](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FContextualAI\u002Fexamples?style=social) | -- |\n| LightRAG | 简单快速的检索增强生成 | [官网](https:\u002F\u002Farxiv.org\u002Fabs\u002F2410.05779) | [Github](https:\u002F\u002Fgithub.com\u002FHKUDS\u002FLightRAG) | [![GitHub](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FHKUDS\u002FLightRAG?color=5B5BD6)](https:\u002F\u002Fgithub.com\u002FHKUDS\u002FLightRAG) | 1天前 |\n| Swarm by OpenAI | 轻量级多智能体编排的教育性框架 | - | [Github](https:\u002F\u002Fgithub.com\u002Fopenai\u002Fswarm) | [![GitHub](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fopenai\u002Fswarm?color=5B5BD6)](https:\u002F\u002Fgithub.com\u002Fopenai\u002Fswarm) | 1天前 |\n| Langroid | 用 Python 轻松构建大语言模型驱动的应用程序的框架 | [官网](https:\u002F\u002Flangroid.github.io\u002Flangroid\u002F) | [Github](https:\u002F\u002Fgithub.com\u002Flangroid\u002Flangroid) | [![GitHub](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Flangroid\u002Flangroid?color=5B5BD6)](https:\u002F\u002Fgithub.com\u002Flangroid\u002Flangroid) | 10小时前 |\n| NeMo-Guardrails | 为基于大语言模型的应用添加可编程护栏 | [官网](https:\u002F\u002Fdocs.nvidia.com\u002Fnemo-guardrails\u002F) | [Github](https:\u002F\u002Fgithub.com\u002FNVIDIA\u002FNeMo-Guardrails) | [![GitHub](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FNVIDIA\u002FNeMo-Guardrails?color=5B5BD6)](https:\u002F\u002Fgithub.com\u002FNVIDIA\u002FNeMo-Guardrails) | 上周 |\n| Swiftide | 用 Rust 构建快速流式大语言模型应用的库 | [官网](https:\u002F\u002Fswiftide.rs) | [GitHub](https:\u002F\u002Fgithub.com\u002Fbosun-ai\u002Fswiftide) | ![GitHub 星数](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fbosun-ai\u002Fswiftide?style=social) | 1小时前 |\n| Korvus | 将整个 RAG 流程整合到一条数据库查询中 | [官网](https:\u002F\u002Fpostgresml.org) | [GitHub](https:\u002F\u002Fgithub.com\u002Fpostgresml\u002Fkorvus) | ![GitHub 星数](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fpostgresml\u002Fkorvus?style=social) | 上月 |\n| semantic-router | 使用语义向量路由大语言模型请求的框架 | [官网](https:\u002F\u002Fwww.aurelio.ai\u002Fsemantic-router) | [GitHub](https:\u002F\u002Fgithub.com\u002Faurelio-labs\u002Fsemantic-router) | ![GitHub 星数](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Faurelio-labs\u002Fsemantic-router?style=social) | 4小时前 |\n| AWS Bedrock Knowledge Bases | 用于构建、扩展和部署 RAG 驱动的应用的服务 | [官网](https:\u002F\u002Faws.amazon.com\u002Fbedrock\u002Fknowledge-bases\u002F) | - | - | 1小时前 |\n| langflow | 构建、扩展和部署 RAG 及多智能体 AI 应用 | [官网](https:\u002F\u002Fwww.langflow.org\u002F) | [Github](https:\u002F\u002Fgithub.com\u002Flangflow-ai\u002Flangflow) | ![GitHub 星数](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Flangflow-ai\u002Flangflow?style=social) | 1小时前 |\n| dspy | 使用模块化编程构建语言模型应用 | [官网](https:\u002F\u002Fdspy-docs.vercel.app\u002F) | [GitHub](https:\u002F\u002Fgithub.com\u002Fstanfordnlp\u002Fdspy) | ![GitHub 星数](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fstanfordnlp\u002Fdspy?style=social) | 13小时前 |\n| mem0 | 您的 AI 应用的记忆层 | [官网](https:\u002F\u002Fmem0.ai\u002F) | [GitHub](https:\u002F\u002Fgithub.com\u002Fmem0ai\u002Fmem0) | ![GitHub 星数](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fmem0ai\u002Fmem0?style=social) | 2小时前 |\n| RAGLite | 用于构建 RAG 应用的 Python 包 | [官网](https:\u002F\u002Fsuperlinear.eu) | [GitHub](https:\u002F\u002Fgithub.com\u002Fsuperlinear-ai\u002Fraglite) | ![GitHub 星数](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fsuperlinear-ai\u002Fraglite?style=social) | 18小时前 |\n| cognee | 用于构建 GraphRAG 应用的记忆框架 | [官网](https:\u002F\u002Fwww.cognee.ai) | [GitHub](https:\u002F\u002Fgithub.com\u002Ftopoteretes\u002Fcognee) | ![GitHub 星数](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Ftopoteretes\u002Fcognee?style=social) | 2小时前 |\n| ragbits | 用于快速开发生成式 AI 应用的构建模块 | [官网](https:\u002F\u002Fragbits.deepsense.ai) | [GitHub](https:\u002F\u002Fgithub.com\u002Fdeepsense-ai\u002Fragbits) | ![GitHub 星数](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fdeepsense-ai\u002Fragbits?style=social) | 1小时前 |\n| Interchange | 从文档上传到搜索的端到端 RAG API | [官网](https:\u002F\u002Fwww.getinterchange.io) | [GitHub](https:\u002F\u002Fgithub.com\u002Fgetinterchange) | [![GitHub](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fgetinterchange?color=5B5BD6)](https:\u002F\u002Fgithub.com\u002Fgetinterchange) | 1小时前 |\n| ZeroEntropy | 重排序器、嵌入及端到端检索 API | [官网](https:\u002F\u002Fdocs.zeroentropy.dev) | [GitHub](https:\u002F\u002Fgithub.com\u002Forgs\u002Fzeroentropy-ai\u002Frepositories) | - | 1小时前 |\n| memori | 多智能体记忆引擎，赋予您的 AI 智能体类人记忆 | [官网](https:\u002F\u002Fmemori.gibsonai.com\u002F) | [GitHub](https:\u002F\u002Fgithub.com\u002Fgibsonai\u002Fmemori) | ![GitHub 星数](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fgibsonai\u002Fmemori?style=social) | 3天前 |\n\n## RAG 评估与优化框架\n\n| 名称 | 描述 | 官网 | GitHub | 星星 | 活跃度 |\n| ------------ | ------------------------------------------------------------------------------------------------------ | ---------------------------------------------------------------------------- | ------------------------------------------------------ | ------------------------------------------------------------------------------------------- | -------- |\n| Trulens | 通过可扩展的评估反馈功能，衡量并提升 LLM 应用的质量 | [官网](https:\u002F\u002Fwww.trulens.org\u002F) | [GitHub](https:\u002F\u002Fgithub.com\u002Ftruera\u002Ftrulens) | ![GitHub 星星](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Ftruera\u002Ftrulens?style=social) | 11小时前 |\n| Phoenix | 面向实验、评估和故障排除设计的 AI 可观测性平台 | [官网](https:\u002F\u002Fgithub.com\u002FArize-ai\u002Fphoenix) | [GitHub](https:\u002F\u002Fgithub.com\u002FArize-ai\u002Fphoenix) | ![GitHub 星星](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FArize-ai\u002Fphoenix?style=social) | 1 天前 |\n| ragas | 评估并量化使用外部数据增强 LLM 上下文的 RAG 流水线性能 | [官网](https:\u002F\u002Fdocs.ragas.io\u002Fen\u002Fstable\u002F) | [GitHub](https:\u002F\u002Fgithub.com\u002Fexplodinggradients\u002Fragas) | ![GitHub 星星](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fexplodinggradients\u002Fragas?style=social) | 3 小时前 |\n| LMUnit | 针对自然语言单元测试评估优化的语言模型 | [官网](https:\u002F\u002Fdocs.contextual.ai\u002Fapi-reference\u002Flmunit\u002Flmunit) | - | - | -- |\n| Deepchecks | 持续验证 AI 和 ML 模型，检测数据漂移和模型问题 | [官网](https:\u002F\u002Fdocs.deepchecks.com\u002Fstable) | [GitHub](https:\u002F\u002Fgithub.com\u002Fdeepchecks\u002Fdeepchecks) | ![GitHub 星星](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fdeepchecks\u002Fdeepchecks?style=social) | 8 分钟前 |\n| AutoRAG | 端到端 RAG 优化：解析、分块、评估数据集创建及流水线部署 | [官网](https:\u002F\u002Fauto-rag.com\u002F) | [GitHub](https:\u002F\u002Fgithub.com\u002FMarker-Inc-Korea\u002FAutoRAG) | ![GitHub 星星](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FMarker-Inc-Korea\u002FAutoRAG?style=social) | 1 小时前 |\n| evalmy.ai | 经过微调的轻量级 RAG 评估服务 + Python 客户端库 | [官网](https:\u002F\u002Fwww.evalmy.ai\u002F) | [GitHub](https:\u002F\u002Fgithub.com\u002Fevalmy-ai\u002Fevalmyai-python) | ![GitHub 星星](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fevalmy-ai\u002Fevalmyai-python?style=social) | -- |\n| TextGrad | 基于 LLM 的文本优化框架，专注于减少幻觉并改进提示语 | [官网](https:\u002F\u002Ftextgrad.com\u002F) | [GitHub](https:\u002F\u002Fgithub.com\u002Fzou-group\u002Ftextgrad) | ![GitHub 星星](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fzou-group\u002Ftextgrad?style=social) | 24 小时前 |\n| langfuse | 跟踪、评估、提示管理及指标，用于调试和改进您的 LLM 应用。 | [官网](https:\u002F\u002Flangfuse.com\u002F) | [GitHub](https:\u002F\u002Fgithub.com\u002Flangfuse\u002Flangfuse) | ![GitHub 星星](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Flangfuse\u002Flangfuse?style=social) | 1 小时前 |\n| Vectara HHEM | 用于 RAG 的幻觉评估模型 | [Huggingface](https:\u002F\u002Fhuggingface.co\u002Fvectara\u002Fhallucination_evaluation_model) | -- | -- | -- |\n| StepsTrack | 一款可观测性工具，用于跟踪、检查和可视化流水线中的每一步 | - | [GitHub](https:\u002F\u002Fgithub.com\u002Flokwkin\u002Fsteps-track) | ![GitHub 星星](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Flokwkin\u002Fsteps-track?style=social) | 15 小时前 |\n| syftr | 多目标端到端代理式 RAG 优化。 | - | [GitHub](https:\u002F\u002Fgithub.com\u002Fdatarobot\u002Fsyftr) | ![GitHub 星星](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fdatarobot\u002Fsyftr?style=social) | 1 小时前 |\n| zbench | 用于检索和重排序的标注与评估框架 | [官网](https:\u002F\u002Fdocs.zeroentropy.dev) | [GitHub](https:\u002F\u002Fgithub.com\u002FZeroEntropy-AI\u002Fzbench) | - | 1 小时前 |\n| rag-select | 跨 RAG 工具产品对端到端 RAG 架构进行评估\u002F优化。 | [官网](https:\u002F\u002Fuseconclude.com\u002Fengineering\u002Frag-select) | [GitHub](https:\u002F\u002Fgithub.com\u002Fconclude-ai\u002Frag-select) | ![GitHub 星星](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fconclude-ai\u002Frag-select?style=social) | 1 小时前 |\n\n## RAG 引擎\n\n| 名称 | 描述 | 官网 | GitHub | 星数 | 活跃度 |\n| -------------------------- | ---------------------------------------------------------------------------------------------- | ----------------------------------------------------------------------------- | ------------------------------------------------------------- | ------------------------------------------------------------------------------------------------ | ------------ |\n| Agentset | 开源的智能体式RAG平台。 | [官网](https:\u002F\u002Fagentset.ai) | [GitHub](https:\u002F\u002Fgithub.com\u002Fagentset-ai\u002Fagentset) | ![GitHub星数](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fagentset-ai\u002Fagentset?style=social) | 1天前 |\n| Engramic | 专注于长期记忆和高级上下文管理的RAG引擎 | [官网](https:\u002F\u002Fengramic.org) | [Github](https:\u002F\u002Fgithub.com\u002Fengramic\u002Fengramic) | ![GitHub星数](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fengramic\u002Fengramic?style=social) | 2小时前 |\n| TrustGraph | 独立于LLM的智能体开发平台 | [官网](https:\u002F\u002Ftrustgraph.ai) | [GitHub](https:\u002F\u002Fgithub.com\u002Ftrustgraph-ai\u002Ftrustgraph) | ![GitHub星数](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Ftrustgraph-ai\u002Ftrustgraph?style=social) | 2天前 |\n| R2R | RAG领域的Elasticsearch，帮助您快速构建和部署可扩展的RAG解决方案 | [官网](https:\u002F\u002Fr2r-docs.sciphi.ai\u002Fintroduction) | [GitHub](https:\u002F\u002Fgithub.com\u002FSciPhi-AI\u002FR2R) | ![GitHub星数](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FSciPhi-AI\u002FR2R?style=social) | 6小时前 |\n| RAGFlow | 基于深度文档理解的开源RAG引擎 | [官网](https:\u002F\u002Fragflow.io) | [GitHub](https:\u002F\u002Fgithub.com\u002Finfiniflow\u002Fragflow) | ![GitHub星数](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Finfiniflow\u002Fragflow?style=social) | 1小时前 |\n| Liquid Index | 统一的RAG平台。一个API。您所需的所有工具 | [官网](https:\u002F\u002Fliquidindex.dev) | - | - | 1小时前 |\n| Vertex AI Knowledge Engine | 用于增强上下文的LLM应用的数据框架 | [官网](https:\u002F\u002Fcloud.google.com\u002Fvertex-ai\u002Fgenerative-ai\u002Fdocs\u002Frag-overview) | - | - | 1天前 |\n| Embedchain | 在不到10行代码内实现LLM响应个性化的开源框架 | [官网](https:\u002F\u002Fdocs.embedchain.ai\u002Fget-started\u002Fquickstart) | [GitHub](https:\u002F\u002Fgithub.com\u002Fmem0ai\u002Fmem0\u002Ftree\u002Fmain\u002Fembedchain) | ![GitHub星数](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fmem0ai\u002Fmem0?style=social) | 上周 |\n| txtai | 用于语义搜索、LLM编排和RAG工作流的一体化嵌入数据库 | [官网](https:\u002F\u002Fneuml.github.io\u002Ftxtai\u002F) | [GitHub](https:\u002F\u002Fgithub.com\u002Fneuml\u002Ftxtai) | ![GitHub星数](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fneuml\u002Ftxtai?style=social) | 上周 |\n| dsRAG | 非结构化数据的高性能检索引擎 | - | [GitHub](https:\u002F\u002Fgithub.com\u002FD-Star-AI\u002FdsRAG) | ![GitHub星数](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FD-Star-AI\u002FdsRAG?style=social) | 上周 |\n| Flash-Rank | 使用成对或列表排序重排器提高检索精度，再传递给LLM。 | - | [GitHub](https:\u002F\u002Fgithub.com\u002FPrithivirajDamodaran\u002FFlashRank) | ![GitHub星数](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FPrithivirajDamodaran\u002FFlashRank?style=social) | 2周前 |\n| Graphlit | 以API优先的平台，用于构建知识驱动的AI应用和智能体 | [官网](https:\u002F\u002Fwww.graphlit.com) | [GitHub](https:\u002F\u002Fgithub.com\u002Fgraphlit) | ![GitHub星数](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fgraphlit?style=social) | 8小时前 |\n| rag-citation | 将RAG与自动引用生成结合，提升内容可信度 | [官网](https:\u002F\u002Fpypi.org\u002Fproject\u002Frag-citation\u002F) | [GitHub](https:\u002F\u002Fgithub.com\u002Frahulanand1103\u002Frag-citation) | ![GitHub星数](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Frahulanand1103\u002Frag-citation?style=social) | 上周 |\n| PostgresML | PostgreSQL + GPU，具备分块、嵌入、转换和排名等功能 | [官网](https:\u002F\u002Fpostgresml.org) | [GitHub](https:\u002F\u002Fgithub.com\u002Fpostgresml\u002Fpostgresml) | ![GitHub星数](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fpostgresml\u002Fpostgresml?style=social) | 昨天 |\n| chainlit | 在几分钟内而非几周内构建生产就绪的对话式AI应用 | [官网](https:\u002F\u002Fchainlit.io\u002F) | [GitHub](https:\u002F\u002Fgithub.com\u002FChainlit\u002Fchainlit) | ![GitHub星数](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FChainlit\u002Fchainlit?style=social) | 24小时前 |\n| pathway | 用于流处理、实时分析、LLM流水线和RAG的Python ETL框架。 | [官网](https:\u002F\u002Fpathway.com\u002F) | [GitHub](https:\u002F\u002Fgithub.com\u002Fpathwaycom\u002Fpathway) | ![GitHub星数](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fpathwaycom\u002Fpathway?style=social) | 7小时前 |\n| cognita | 用于模块化、开源生产级应用的RAG框架。 | [官网](https:\u002F\u002Fcognita.truefoundry.com\u002F) | [GitHub](https:\u002F\u002Fgithub.com\u002Ftruefoundry\u002Fcognita) | ![GitHub星数](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Ftruefoundry\u002Fcognita?style=social) | 2天前 |\n| FlashRAG | 用于高效RAG研究的Python工具包 | - | [GitHub](https:\u002F\u002Fgithub.com\u002FRUC-NLPIR\u002FFlashRAG) | ![GitHub星数](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FRUC-NLPIR\u002FFlashRAG?style=social) | 3小时前 |\n| RAGatouille | 轻松在任何RAG管道中训练和使用先进的检索方法。 | - | [GitHub](https:\u002F\u002Fgithub.com\u002FAnswerDotAI\u002FRAGatouille) | ![GitHub星数](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FAnswerDotAI\u002FRAGatouille?style=social) | 4个月前 |\n| pgai | 一套仅在PostgreSQL中即可开发RAG、语义搜索及其他AI应用的工具套件 | [官网](https:\u002F\u002Fwww.timescale.com\u002Fai) | [GitHub](https:\u002F\u002Fgithub.com\u002Ftimescale\u002Fpgai) | ![GitHub星数](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Ftimescale\u002Fpgai?style=social) | 10小时前 |\n| Vectara | 可信赖的RAG平台，用于快速构建AI助手和智能体。 | [官网](https:\u002F\u002Fwww.vectara.com\u002F) | [GitHub](https:\u002F\u002Fgithub.com\u002Fvectara\u002F) | - | - |\n| mode | 具有专家模型、智能聚类和小数据集高效检索功能的RAG框架。 | - | [GitHub](https:\u002F\u002Fgithub.com\u002Frahulanand1103\u002Fmode) |![GitHub星数](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Frahulanand1103\u002Fmode?style=social) | 2天前 |\n| haiku.rag | 开源RAG框架，支持监控、CLI、搜索、问答、MCP，并可在SQLite上运行。 | - | [Github](https:\u002F\u002Fgithub.com\u002Fggozad\u002Fhaiku.rag) |![GitHub星数](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fggozad\u002Fhaiku.rag?style=social) | 3小时前 |\n| ZeroEntropy AI | 开放权重的重排序器、嵌入及端到端检索API | [官网](https:\u002F\u002Fdocs.zeroentropy.dev) | [Github](https:\u002F\u002Fgithub.com\u002FZeroEntropy-AI) | - | 3小时前 |\n\n## RAG 数据准备框架\n\n| 名称 | 描述 | 官网 | GitHub | 星标 | 活跃度 |\n| --------- | ---------------------------------- | -------------------------------- | --------------------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------- | -------- |\n| CocoIndex | 用于构建全新索引的 ETL 框架 | [官网](https:\u002F\u002Fcocoindex.io\u002F) | [Github](https:\u002F\u002Fgithub.com\u002Fcocoindex-io\u002Fcocoindex) | [![GitHub](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fcocoindex-io\u002Fcocoindex?color=5B5BD6)](https:\u002F\u002Fgithub.com\u002Fcocoindex-io\u002Fcocoindex) | 1 小时前 |\n| Gitana.io | 面向编辑审批的内容平台，支持将训练好的数据集定时部署到 RAG 向量数据库 | [官网](https:\u002F\u002Fgitana.io\u002F) | - | - | - |\n| Chonkie | 简洁、轻量且快速的 RAG 分块库 | [官网](https:\u002F\u002Fchonkie.ai\u002F) | [GitHub](https:\u002F\u002Fgithub.com\u002Fchonkie-inc\u002Fchonkie) | [![GitHub 星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fchonkie-inc\u002Fchonkie?style=social)](https:\u002F\u002Fgithub.com\u002Fchonkie-inc\u002Fchonkie) | 1 小时前 |\n\n## RAG 项目\n\n| 名称 | 描述 | 官网 | GitHub | 星数 | 活跃度 |\n|------------------------------------------|--------------------------------------------------------------------------------------------------------------------| ----------------------------------------------------------------------------------------- | ----------------------------------- | ------------------------- | --------- |\n| LlamaParse | 原生支持生成式AI的文档解析平台 | [官网](https:\u002F\u002Fdocs.llamaindex.ai\u002Fen\u002Fstable\u002Fllama_cloud\u002Fllama_parse\u002F) | [GitHub](https:\u002F\u002Fgithub.com\u002Frun-llama\u002Fllama_parse) | ![GitHub星数](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Frun-llama\u002Fllama_parse?style=social) | 2天前 |\n| Langchain-extract | 使用大语言模型从文本和文件中提取信息的Web服务器 | [官网](https:\u002F\u002Fpython.langchain.com\u002Fv0.1\u002Fdocs\u002Fuse_cases\u002Fextraction\u002F) | [GitHub](https:\u002F\u002Fgithub.com\u002Flangchain-ai\u002Flangchain-extract) | ![GitHub星数](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Flangchain-ai\u002Flangchain-extract?style=social) | 4分钟前 |\n| Needle | 开箱即用的生产级RAG流水线。 | [官网](https:\u002F\u002Fneedle-ai.com\u002F) | [GitHub](https:\u002F\u002Fgithub.com\u002Foeken\u002Fneedle-python) | ![GitHub星数](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Foeken\u002Fneedle-python?style=social) | 1小时前 |\n| Unstructured.io | 构建用于标注、训练或生产环境的自定义预处理流水线 | [官网](https:\u002F\u002Funstructured.io\u002F) | [GitHub](https:\u002F\u002Fgithub.com\u002FUnstructured-IO\u002Funstructured) | ![GitHub星数](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FUnstructured-IO\u002Funstructured?style=social) | 3天前 |\n| Verba | 由Weaviate驱动的RAG聊天机器人 | [官网](https:\u002F\u002Fverba.weaviate.io\u002F) | [GitHub](https:\u002F\u002Fgithub.com\u002Fweaviate\u002FVerba) | ![GitHub星数](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fweaviate\u002FVerba?style=social) | 2周前 |\n| Unstract | 无代码平台，用于部署API和ETL管道以结构化非结构化文档 | [官网](https:\u002F\u002Funstract.com\u002F) | [GitHub](https:\u002F\u002Fgithub.com\u002FZipstack\u002Funstract) | ![GitHub星数](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FZipstack\u002Funstract?style=social) | 4小时前 |\n| Humata.ai | 跨所有文档文件提问 | [官网](https:\u002F\u002Fwww.humata.ai\u002F) | - | - | 4小时前 |\n| Kiln | 构建包含评估、RAG、智能体、微调和合成数据的AI系统 | [官网](https:\u002F\u002Fkiln.tech) | [GitHub](https:\u002F\u002Fgithub.com\u002FKiln-AI\u002FKiln) | ![GitHub星数](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FKiln-AI\u002FKiln?style=social) | 12小时前 |\n| Ragie.ai | 为开发者提供的完全托管的RAG即服务。 | [官网](https:\u002F\u002Fwww.ragie.ai\u002F) | [GitHub](https:\u002F\u002Fgithub.com\u002Fragieai) | - | 12小时前 |\n| Reducto | 解析复杂文档并创建适合LLM使用的输入 | [官网](https:\u002F\u002Freducto.ai\u002F) | [GitHub](https:\u002F\u002Fgithub.com\u002Freductoai) | - | 2周前 |\n| Midship | 直接将文档数据提取到您的电子表格\u002FERP\u002FCRM | [官网](https:\u002F\u002Fmidship.ai\u002F) | - | - | - |\n| DocuPanda | 将文档转换为结构化、标准化的字段和值集合 | [官网](https:\u002F\u002Fwww.docupanda.io\u002F) | - | - | - |\n| contextual-doc-retrieval-opneai-reranker | 使用GPT-4和Cohere进行查询扩展，并结合BM25进行重排序 | - | [GitHub](https:\u002F\u002Fgithub.com\u002Flesteroliver911\u002Fcontextual-doc-retrieval-opneai-reranker) | ![GitHub星数](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Flesteroliver911\u002Fcontextual-doc-retrieval-opneai-reranker?style=social) | 上周 |\n| Raggenie | 低代码平台，用于构建自定义的基于RAG的AI应用 | [官网](https:\u002F\u002Fwww.raggenie.com) | [GitHub](https:\u002F\u002Fgithub.com\u002Fsirocco-ventures\u002Fraggenie) | ![GitHub星数](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fsirocco-ventures\u002Fraggenie?style=social) | 10小时前 |\n| Chunkr | 基于视觉模型的PDF分块和OCR，专为快速处理大规模数据集优化 | [官网](https:\u002F\u002Fchunkr.ai) | [GitHub](https:\u002F\u002Fgithub.com\u002Flumina-ai-inc\u002Fchunkr) | ![GitHub星数](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Flumina-ai-inc\u002Fchunkr?style=social) | 11小时前 |\n| tldw | 类似于NotebookLM的开源项目 | [官网](https:\u002F\u002Ftldwproject.com) | [GitHub](https:\u002F\u002Fgithub.com\u002Frmusser01\u002Ftldw) | ![GitHub星数](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Frmusser01\u002Ftldw?style=social) | 昨天 |\n| Cerbos | 面向RAG和LLM的访问控制。 | [官网](https:\u002F\u002Fsolutions.cerbos.dev\u002Fauthorization-in-rag-based-ai-systems-with-cerbos) | [GitHub](https:\u002F\u002Fgithub.com\u002Fcerbos\u002Fcerbos) | ![GitHub星数](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fcerbos\u002Fcerbos?style=social) | 14小时前 |\n| extractous | 为您的AI应用提供极快的数据提取 | [官网](https:\u002F\u002Fwww.extractous.com\u002F) | [GitHub](https:\u002F\u002Fgithub.com\u002Fyobix-ai\u002Fextractous) | ![GitHub星数](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fyobix-ai\u002Fextractous?style=social) | - |\n| SWIRL | 面向工作场所的AI搜索与RAG。即时获取公司知识库中的AI洞察。 | [官网](https:\u002F\u002Fwww.swirlaiconnect.com\u002F) | [GitHub](https:\u002F\u002Fgithub.com\u002Fswirlai\u002Fswirl-search) | ![GitHub星数](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fswirlai\u002Fswirl-search?style=social) | 2周前 |\n| ChatDOC PDF Parser | 高精度PDF解析，可将文档转化为适用于RAG系统的完美结构化数据。 | [官网](https:\u002F\u002Fpdfparser.io\u002F?src=github) | - | - | - |\n| Gurubase | 通过索引网站、PDF文档、YouTube视频和GitHub代码仓库，创建AI驱动的问答助手。 | [官网](https:\u002F\u002Fgurubase.io) | [GitHub](https:\u002F\u002Fgithub.com\u002FGurubase\u002Fgurubase) | ![GitHub星数](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FGurubase\u002Fgurubase?style=social) | 1天前 |\n| Archive Agent | 开源语义文件追踪器，具备OCR + AI搜索功能。内置RAG引擎的智能索引器。 | - | [GitHub](https:\u002F\u002Fgithub.com\u002FshredEngineer\u002FArchive-Agent) | ![GitHub星数](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FshredEngineer\u002FArchive-Agent?style=social) | - |\n| MidrasAI | Colpali多模态检索模型的简单API。 | - | [Github](https:\u002F\u002Fgithub.com\u002Fajac-zero\u002Fmidrasai) | ![GitHub星数](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fajac-zero\u002Fmidrasai?style=social) | 6分钟前 |\n| EmbeddingBridge | 用于嵌入的版本控制和迁移工具 | - | [Github](https:\u002F\u002Fgithub.com\u002FProgramComputer\u002FEmbeddingBridge) | ![Github星数](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FProgramComputer\u002FEmbeddingBridge?style=social)|-|\n| Stream-Rag-Agent | 面向Kafka的流式RAG代理 | - | [Github](https:\u002F\u002Fgithub.com\u002Fonurbaran\u002Fstream-rag-agent) | |-|\n| zchunk | 开源高效的基于LLM的分块方法 | [官网](https:\u002F\u002Fdocs.zeroentropy.dev) | [Github](https:\u002F\u002Fgithub.com\u002FZeroEntropy-AI\u002Fzchunk) | - | 2小时前\n| hydrot | 一款面向企业文档的生产就绪型检索增强生成（RAG）系统，对Markdown内容提供一流的支持。采用微服务架构，实现水平扩展和灵活性。 | - | [Github](https:\u002F\u002Fgithub.com\u002Frjxby\u002Fhydrot) | ![Github星数](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Frjxby\u002Fhydrot?style=social)|-|\n| Tensorlake | 面向可引用、可追溯的RAG系统的文档解析。提取文本、表格和图像，同时保持复杂的布局，附带边界框坐标，并支持结构化输出 | [官网](https:\u002F\u002Fwww.tensorlake.ai\u002F) | [Github](https:\u002F\u002Fgithub.com\u002Ftensorlakeai) | ![Github星数](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Ftensorlakeai\u002Findexify?style=social) | 2天前|\n\n## RAG 资源与站点\n\n| 站点\u002F文章 | 描述 | 链接 |\n| -------------------- | ----------------------------------------------------------------------------- | -------------------------------------------------------------- |\n| 上下文检索 | Anthropic 推出上下文检索功能 | [官网](https:\u002F\u002Fwww.anthropic.com\u002Fnews\u002Fcontextual-retrieval) |\n| Open-RAG | 基于开源大语言模型的增强型检索增强推理 | [官网](https:\u002F\u002Farxiv.org\u002Fabs\u002F2410.01782) |\n| ColPali | 使用视觉语言模型实现高效文档检索 | [官网](https:\u002F\u002Farxiv.org\u002Fabs\u002F2407.01449) |\n| RAG_Techniques | 展示多种 RAG 系统的高级技术 | [官网](https:\u002F\u002Fgithub.com\u002FNirDiamant\u002FRAG_Techniques) |\n| GenAI_Agents | 各类 AI 代理技术的教程与实现 | [官网](https:\u002F\u002Fgithub.com\u002FNirDiamant\u002FGenAI_Agents) |\n| RAG From Scratch | LangChain 官方指南：从零开始构建 RAG 流程 | [官网](https:\u002F\u002Fgithub.com\u002Flangchain-ai\u002Frag-from-scratch) |\n\n\n## 模型排行榜\n\n| 名称 | 描述 | 链接 |\n| --------------------------------- | --------------------------- | --------------------------------------------------------------- |\n| Artificial Analysis | 大语言模型比较 | [官网](https:\u002F\u002Fartificialanalysis.ai) |\n| HuggingFace\u002Fmteb | 嵌入模型排行榜 | [官网](https:\u002F\u002Fhuggingface.co\u002Fspaces\u002Fmteb\u002Fleaderboard) |\n| Vectara 幻觉排行榜 | 大语言模型幻觉排行榜 | [官网](https:\u002F\u002Fgithub.com\u002Fvectara\u002Fhallucination-leaderboard) |\n\n> 如果您正在寻找主流的 RAG 框架和技巧，请查看 Nir Diamant 的优秀仓库：[RAG Techniques](https:\u002F\u002Fgithub.com\u002FNirDiamant\u002FRAG_Techniques)。该仓库专注于已在社区中获得广泛认可的成熟工具和方法。\n\n## 许可证\n\n本项目采用 **MIT 许可证**。详情请参阅 [LICENSE](LICENSE) 文件。\n\n## 加入讨论\n\n本项目是 [r\u002FRAG](https:\u002F\u002Fwww.reddit.com\u002Fr\u002FRag\u002F) 社区的一部分。您有任何反馈或建议吗？欢迎随时提交问题、发起讨论，或加入我们的 [Discord 服务器](https:\u002F\u002Fdiscord.gg\u002Fnn92wC5QmN)！我们希望将此仓库打造成为所有探索 RAG 生态系统的用户的宝贵资源，您的意见对我们至关重要。","# RAGHub 快速上手指南\n\n**注意**：RAGHub 本身不是一个单一的软件开发工具包（SDK）或可安装的软件库，而是一个**检索增强生成（RAG）生态系统的工具目录和资源列表**。它旨在帮助开发者发现、比较和选择适合的 RAG 框架、引擎及评估工具。\n\n因此，本指南将指导你如何**浏览该目录**并**快速启动其中推荐的主流 RAG 框架**（以 LangChain 和 LlamaIndex 为例），以便你立即开始构建应用。\n\n## 1. 环境准备\n\n在开始使用 RAGHub 列出的任何工具之前，请确保你的开发环境满足以下通用要求：\n\n* **操作系统**：Linux, macOS, 或 Windows (WSL2 推荐)\n* **Python 版本**：建议安装 Python 3.9 或更高版本\n* **包管理器**：`pip` 或 `conda`\n* **前置依赖**：\n * Git (用于克隆项目代码)\n * 基础深度学习库 (如 `torch`, `transformers`，通常由具体框架自动安装)\n\n**国内加速建议**：\n推荐使用国内镜像源加速 Python 包的安装，例如清华大学或阿里云镜像。\n```bash\npip config set global.index-url https:\u002F\u002Fpypi.tuna.tsinghua.edu.cn\u002Fsimple\n```\n\n## 2. 安装步骤\n\n由于 RAGHub 是目录，你需要根据需求选择具体的框架进行安装。以下是两个最流行框架的安装命令：\n\n### 选项 A：安装 LangChain (通用型框架)\n适合需要灵活编排链式调用和集成多种模型的场景。\n\n```bash\npip install langchain langchain-community langchain-core\n# 如需向量数据库支持，可选装\npip install chromadb\n```\n\n### 选项 B：安装 LlamaIndex (数据连接型框架)\n适合专注于数据索引、结构化查询和复杂数据检索的场景。\n\n```bash\npip install llama-index\n```\n\n### 选项 C：浏览完整目录\n你可以直接访问 RAGHub 的在线仓库或本地克隆以查看最新工具列表：\n\n```bash\ngit clone https:\u002F\u002Fgithub.com\u002FContextualAI\u002FRAGHub.git\ncd RAGHub\n# 查看 README.md 获取完整工具列表\n```\n\n## 3. 基本使用\n\n以下展示如何使用上述安装的框架运行一个最简单的 \"Hello World\" 级别的 RAG 流程（基于本地文本的检索与生成）。\n\n### 示例 1：使用 LangChain 构建简单 RAG\n\n```python\nfrom langchain.text_splitter import CharacterTextSplitter\nfrom langchain_community.document_loaders import TextLoader\nfrom langchain_community.embeddings import HuggingFaceEmbeddings\nfrom langchain_community.vectorstores import Chroma\nfrom langchain.chains import RetrievalQA\nfrom langchain_community.llms import HuggingFacePipeline\nfrom transformers import AutoTokenizer, AutoModelForCausalLM, pipeline\n\n# 1. 准备数据\ntext = \"RAGHub 是一个用于检索增强生成 (RAG) 的工具目录。它帮助开发者找到最新的框架。\"\nloader = TextLoader.from_texts([text])\ndocuments = loader.load()\n\n# 2. 分割文本\ntext_splitter = CharacterTextSplitter(chunk_size=50, chunk_overlap=0)\ndocs = text_splitter.split_documents(documents)\n\n# 3. 创建嵌入和向量存储 (使用本地模型)\nembeddings = HuggingFaceEmbeddings(model_name=\"sentence-transformers\u002Fall-MiniLM-L6-v2\")\nvectorstore = Chroma.from_documents(docs, embeddings)\n\n# 4. 设置 LLM (此处以占位符为例，实际需加载具体模型)\n# llm = HuggingFacePipeline(...) \n\n# 5. 构建检索链\n# qa_chain = RetrievalQA.from_chain_type(llm, retriever=vectorstore.as_retriever())\n\n# 6. 提问\n# result = qa_chain.run(\"RAGHub 是什么？\")\n# print(result)\nprint(\"LangChain 环境已就绪，可开始构建 RAG 应用。\")\n```\n\n### 示例 2：使用 LlamaIndex 构建简单 RAG\n\n```python\nfrom llama_index.core import SimpleDirectoryReader, VectorStoreIndex, Settings\nfrom llama_index.embeddings.huggingface import HuggingFaceEmbedding\n\n# 1. 设置嵌入模型\nSettings.embed_model = HuggingFaceEmbedding(model_name=\"BAAI\u002Fbge-small-zh-v1.5\")\n\n# 2. 加载文档 (假设当前目录下有 data.txt)\n# documents = SimpleDirectoryReader(\".\u002F\").load_data()\n# 模拟文档加载\nfrom llama_index.core import Document\ndocuments = [Document(text=\"RAGHub 是一个用于检索增强生成 (RAG) 的工具目录。\")]\n\n# 3. 构建索引\nindex = VectorStoreIndex.from_documents(documents)\n\n# 4. 创建查询引擎\nquery_engine = index.as_query_engine()\n\n# 5. 查询\n# response = query_engine.query(\"RAGHub 是做什么的？\")\n# print(response)\nprint(\"LlamaIndex 环境已就绪，可开始构建 RAG 应用。\")\n```\n\n**下一步建议**：\n回到 RAGHub 目录，根据具体需求（如：需要评估框架、特定的 RAG 引擎或数据预处理工具）选择其他列出的项目进行深入探索。","某初创团队正急于构建一款基于内部技术文档的智能客服机器人，需要在两周内完成从选型到原型上线的全流程。\n\n### 没有 RAGHub 时\n- **选型如大海捞针**：面对 GitHub 上成千上万个 RAG 项目，开发人员花费数天搜索关键词，却难以区分哪些是三个月前的过时方案，哪些是刚发布的创新框架。\n- **陷入“重复造轮子”陷阱**：因缺乏权威目录参考，团队误选了一个仅包装旧概念的新工具，导致开发中途发现功能缺失，被迫推倒重来。\n- **评估成本极高**：为了验证框架的活跃度和社区支持，工程师需手动检查每个项目的 Issue 关闭率、最后提交时间和 Star 增长趋势，效率极低。\n- **资源分散难整合**：数据预处理、向量检索引擎和评估工具散落在不同论坛和博客中，缺乏系统化的分类指引，导致技术栈拼接困难。\n\n### 使用 RAGHub 后\n- **精准锁定前沿工具**：通过 RAGHub 的分类目录，团队迅速筛选出\"1 小时前”仍有更新的 Dcup 和 LangChain 等高活跃度框架，直接锁定最新技术栈。\n- **规避炒作与过时风险**：借助社区驱动的实时维护机制，团队清晰识别出哪些项目是真正的创新，哪些只是换皮炒作，避免了无效投入。\n- **一键获取关键指标**：RAGHub 直接展示各框架的 Star 数、最后活动时间及官方链接，让技术决策从“凭感觉”变为“看数据”，选型时间缩短 80%。\n- **全链路资源一站式配齐**：从数据准备框架到评估优化方案，RAGHub 提供了完整的生态地图，团队得以快速组合出最优解决方案并立即启动开发。\n\nRAGHub 将混乱的 RAG 生态转化为清晰的导航图，帮助开发者在技术爆炸时代以最低成本做出最明智的架构决策。","https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002FAndrew-Jang_RAGHub_ae75053a.png","Andrew-Jang","Andrew Jang","https:\u002F\u002Foss.gittoolsai.com\u002Favatars\u002FAndrew-Jang_d3c0920a.png","Building Chalcack",null,"Seoul, South Korea","https:\u002F\u002Fgithub.com\u002FAndrew-Jang",1705,155,"2026-04-05T05:13:08","MIT",1,"","未说明",{"notes":90,"python":88,"dependencies":91},"RAGHub 本身不是一个可运行的软件工具，而是一个收录了多种 RAG（检索增强生成）框架、项目和资源的目录列表。因此，它没有统一的运行环境、依赖库或硬件需求。具体的环境要求取决于用户选择使用的列表中的某个特定工具（如 LangChain, LlamaIndex, Haystack 等）。",[],[26,14,13,54,15],[94,95,96,97,98,99,100,101,102,103],"ai","artificial-intelligence","large-language-models","llm","machine-learning","natural-language-processing","nlp","open-source","rag","retrieval-augmented-generation","2026-03-27T02:49:30.150509","2026-04-06T05:36:32.022222",[107,112,117,122,126,131],{"id":108,"question_zh":109,"answer_zh":110,"source_url":111},18221,"如何向 RAGHub 列表添加新的工具或项目？","请创建一个新的分支（branch），在对应的分类下添加您的项目信息，然后提交合并请求（Pull Request）。维护者审核后会批准合并。具体步骤请参考仓库中的 CONTRIBUTING.md 文件。","https:\u002F\u002Fgithub.com\u002FAndrew-Jang\u002FRAGHub\u002Fissues\u002F71",{"id":113,"question_zh":114,"answer_zh":115,"source_url":116},18222,"提交新项目时应该放在哪个分类下？","请将新项目添加在您希望用户看到它的特定分类下。例如，RAG 框架类项目通常归入\"RAG Projects\"类别。未来计划增加\"Document\u002FFiles\"（文档\u002F文件）类别。如果不确定，可以参考现有项目的分类方式或查看 CONTRIBUTING.md。","https:\u002F\u002Fgithub.com\u002FAndrew-Jang\u002FRAGHub\u002Fissues\u002F36",{"id":118,"question_zh":119,"answer_zh":120,"source_url":121},18223,"我想推荐一个 RAG 相关的库，是直接提 Issue 就可以吗？","不建议直接在 Issue 中等待添加。正确的流程是：阅读 CONTRIBUTING.md 指南，自行创建分支并修改列表文件，然后发起 Pull Request（PR）。维护者会在 PR 中进行审核和合并，这样效率更高。","https:\u002F\u002Fgithub.com\u002FAndrew-Jang\u002FRAGHub\u002Fissues\u002F56",{"id":123,"question_zh":124,"answer_zh":125,"source_url":116},18224,"对于高星级的 RAG 项目（如 kotaemon），如何确定其分类？","对于知名的 RAG 框架项目，可以直接归类为\"RAG Projects\"。如果您发现某个项目适合但尚未收录，可以按照标准流程（新建分支并提交 PR）将其添加到列表中，无需过度犹豫。",{"id":127,"question_zh":128,"answer_zh":129,"source_url":130},18225,"提交添加工具的请求后，维护者通常会如何回应？","维护者通常会回复\"请创建新分支并请求合并，我会批准\"。这意味着您需要主动发起 Pull Request 而不是仅停留在 Issue 讨论阶段。一旦 PR 提交，维护者会尽快进行代码审查和合并操作。","https:\u002F\u002Fgithub.com\u002FAndrew-Jang\u002FRAGHub\u002Fissues\u002F55",{"id":132,"question_zh":133,"answer_zh":134,"source_url":111},18226,"在哪里可以找到贡献指南和具体的添加步骤？","所有贡献步骤和格式要求都详细记录在仓库根目录的 CONTRIBUTING.md 文件中。在提交任何内容之前，请务必仔细阅读该文件，以确保您的提交符合项目规范，从而提高合并成功率。",[]]