[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"similar-ilsilfverskiold--Awesome-LLM-Resources-List":3,"tool-ilsilfverskiold--Awesome-LLM-Resources-List":61},[4,18,28,36,45,53],{"id":5,"name":6,"github_repo":7,"description_zh":8,"stars":9,"difficulty_score":10,"last_commit_at":11,"category_tags":12,"status":17},4358,"openclaw","openclaw\u002Fopenclaw","OpenClaw 是一款专为个人打造的本地化 AI 助手，旨在让你在自己的设备上拥有完全可控的智能伙伴。它打破了传统 AI 助手局限于特定网页或应用的束缚，能够直接接入你日常使用的各类通讯渠道，包括微信、WhatsApp、Telegram、Discord、iMessage 等数十种平台。无论你在哪个聊天软件中发送消息，OpenClaw 都能即时响应，甚至支持在 macOS、iOS 和 Android 设备上进行语音交互，并提供实时的画布渲染功能供你操控。\n\n这款工具主要解决了用户对数据隐私、响应速度以及“始终在线”体验的需求。通过将 AI 部署在本地，用户无需依赖云端服务即可享受快速、私密的智能辅助，真正实现了“你的数据，你做主”。其独特的技术亮点在于强大的网关架构，将控制平面与核心助手分离，确保跨平台通信的流畅性与扩展性。\n\nOpenClaw 非常适合希望构建个性化工作流的技术爱好者、开发者，以及注重隐私保护且不愿被单一生态绑定的普通用户。只要具备基础的终端操作能力（支持 macOS、Linux 及 Windows WSL2），即可通过简单的命令行引导完成部署。如果你渴望拥有一个懂你",349277,3,"2026-04-06T06:32:30",[13,14,15,16],"Agent","开发框架","图像","数据工具","ready",{"id":19,"name":20,"github_repo":21,"description_zh":22,"stars":23,"difficulty_score":24,"last_commit_at":25,"category_tags":26,"status":17},9989,"n8n","n8n-io\u002Fn8n","n8n 是一款面向技术团队的公平代码（fair-code）工作流自动化平台，旨在让用户在享受低代码快速构建便利的同时，保留编写自定义代码的灵活性。它主要解决了传统自动化工具要么过于封闭难以扩展、要么完全依赖手写代码效率低下的痛点，帮助用户轻松连接 400 多种应用与服务，实现复杂业务流程的自动化。\n\nn8n 特别适合开发者、工程师以及具备一定技术背景的业务人员使用。其核心亮点在于“按需编码”：既可以通过直观的可视化界面拖拽节点搭建流程，也能随时插入 JavaScript 或 Python 代码、调用 npm 包来处理复杂逻辑。此外，n8n 原生集成了基于 LangChain 的 AI 能力，支持用户利用自有数据和模型构建智能体工作流。在部署方面，n8n 提供极高的自由度，支持完全自托管以保障数据隐私和控制权，也提供云端服务选项。凭借活跃的社区生态和数百个现成模板，n8n 让构建强大且可控的自动化系统变得简单高效。",184740,2,"2026-04-19T23:22:26",[16,14,13,15,27],"插件",{"id":29,"name":30,"github_repo":31,"description_zh":32,"stars":33,"difficulty_score":10,"last_commit_at":34,"category_tags":35,"status":17},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,"2026-04-05T11:01:52",[14,15,13],{"id":37,"name":38,"github_repo":39,"description_zh":40,"stars":41,"difficulty_score":24,"last_commit_at":42,"category_tags":43,"status":17},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",161147,"2026-04-19T23:31:47",[14,13,44],"语言模型",{"id":46,"name":47,"github_repo":48,"description_zh":49,"stars":50,"difficulty_score":24,"last_commit_at":51,"category_tags":52,"status":17},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。",109154,"2026-04-18T11:18:24",[14,15,13],{"id":54,"name":55,"github_repo":56,"description_zh":57,"stars":58,"difficulty_score":24,"last_commit_at":59,"category_tags":60,"status":17},6121,"gemini-cli","google-gemini\u002Fgemini-cli","gemini-cli 是一款由谷歌推出的开源 AI 命令行工具，它将强大的 Gemini 大模型能力直接集成到用户的终端环境中。对于习惯在命令行工作的开发者而言，它提供了一条从输入提示词到获取模型响应的最短路径，无需切换窗口即可享受智能辅助。\n\n这款工具主要解决了开发过程中频繁上下文切换的痛点，让用户能在熟悉的终端界面内直接完成代码理解、生成、调试以及自动化运维任务。无论是查询大型代码库、根据草图生成应用，还是执行复杂的 Git 操作，gemini-cli 都能通过自然语言指令高效处理。\n\n它特别适合广大软件工程师、DevOps 人员及技术研究人员使用。其核心亮点包括支持高达 100 万 token 的超长上下文窗口，具备出色的逻辑推理能力；内置 Google 搜索、文件操作及 Shell 命令执行等实用工具；更独特的是，它支持 MCP（模型上下文协议），允许用户灵活扩展自定义集成，连接如图像生成等外部能力。此外，个人谷歌账号即可享受免费的额度支持，且项目基于 Apache 2.0 协议完全开源，是提升终端工作效率的理想助手。",100752,"2026-04-10T01:20:03",[27,13,15,14],{"id":62,"github_repo":63,"name":64,"description_en":65,"description_zh":66,"ai_summary_zh":66,"readme_en":67,"readme_zh":68,"quickstart_zh":69,"use_case_zh":70,"hero_image_url":71,"owner_login":72,"owner_name":73,"owner_avatar_url":74,"owner_bio":75,"owner_company":75,"owner_location":76,"owner_email":77,"owner_twitter":75,"owner_website":75,"owner_url":78,"languages":79,"stars":84,"forks":85,"last_commit_at":86,"license":75,"difficulty_score":87,"env_os":88,"env_gpu":89,"env_ram":89,"env_deps":90,"category_tags":93,"github_topics":94,"view_count":24,"oss_zip_url":75,"oss_zip_packed_at":75,"status":17,"created_at":98,"updated_at":99,"faqs":100,"releases":101},9936,"ilsilfverskiold\u002FAwesome-LLM-Resources-List","Awesome-LLM-Resources-List","A Curated Collection of resources for applied AI engineering (work in progress).","Awesome-LLM-Resources-List 是一个专为应用级 AI 工程打造的精选资源库，旨在帮助开发者高效落地大语言模型（LLM）项目。面对市面上繁杂的模型部署方案与云服务选项，它系统地梳理了各类“无服务器”托管平台（如 Modal、Baseten、Hugging Face Endpoints 等），通过直观的对比表格，清晰展示了各工具在冷启动速度、开源模型支持度、开发体验及免费额度等关键维度的差异。\n\n此外，该资源库还深入分析了不同平台在计算定价、请求限制、超时阈值及并发能力上的具体细节，为用户提供了极具参考价值的成本与性能评估数据。这有效解决了技术选型难、环境配置复杂以及成本预估不透明等痛点，让团队能快速找到最适合自身需求的部署方案。\n\n无论是正在探索模型落地的 AI 工程师、需要控制预算的研究人员，还是希望快速构建原型的初创团队，都能从中获益。其独特的亮点在于不仅罗列资源，更提供了横向的技术指标对比和实战指南，将复杂的云基础设施决策过程变得简单透明，是连接大模型技术与实际工程应用的实用桥梁。","# 🌟 Awesome LLM Resources\n\nA Curated Collection of LLM resources. 💡✨\n\n**🌐 Updated: 22nd of June 2025**\n\n### 'Serverless' Hosting of Private\u002FOS Models\n\n| Platform\u002FTool | Rel. | Scale Down | OS 🔓 | GH | Start | One-Click | Dev Exp. | Free-Tier |\n| --------------------------------| -------- | -------------| -------------- | ----------|----------------|---------------|-------------------------|--------------|\n| [Baseten](https:\u002F\u002Fwww.baseten.com\u002F) | 2019 | > 15 min | 🔴 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fbasetenlabs?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fbasetenlabs) | [Guide](https:\u002F\u002Fdocs.baseten.co\u002Fdeploy\u002Fguides\u002Fprivate-model) | 🟡 | 👍 | $30 |\n| [Modal](https:\u002F\u002Fmodal.com\u002F) | 2021 | \u003C 1 min | 🔴 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fmodal-labs?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fmodal-labs) | [Helpers](https:\u002F\u002Fgithub.com\u002Filsilfverskiold\u002FAwesome-LLM-Resources-List\u002Ftree\u002Fmain\u002Fhelpers\u002Fscripts\u002Fmodal) | ❌ | 👍 | $30\u002Fm |\n| [HF Endpoints](https:\u002F\u002Fui.endpoints.huggingface.co\u002F) | 2023 | > 15 min | 🔴 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fhuggingface?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fhuggingface) | None Needed | ✅ | 😓 | ❌ |\n| [Replicate](https:\u002F\u002Freplicate.com\u002F) | 2019 | \u003C 1 min | 🔴 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Freplicate?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Freplicate) | [Guide](https:\u002F\u002Freplicate.com\u002Fdocs\u002Fguides\u002Fpush-a-transformers-model) | 🟡 | 🤷 | ❌ |\n| [Sagemaker (Serverless)](https:\u002F\u002Faws.amazon.com\u002Fsagemaker\u002F) | 2017 | N\u002FA | 🔴 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Faws?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Faws\u002Famazon-sagemaker-examples) | N\u002FA | ❌ | ❌ | 300,000s |\n| [Lambda w\u002F EFS (AWS)](https:\u002F\u002Faws.amazon.com\u002Fpm\u002Flambda\u002F) | 2014 | \u003C 1 min | 🔴 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fawsdocs?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fawsdocs\u002Faws-lambda-developer-guide) | [Guide](https:\u002F\u002Faws.amazon.com\u002Fblogs\u002Fcompute\u002Fhosting-hugging-face-models-on-aws-lambda\u002F) | ❌ | ❌ | ✅ |\n| [RunPod Serverless](https:\u002F\u002Fwww.runpod.io\u002Fserverless-gpu) | 2022 | > 30s | 🔴 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Frunpod?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Frunpod) | N\u002FA | ❌ | 🤷 | ❌ |\n| [BentoML](https:\u002F\u002Fwww.bentoml.com\u002F) | 2019 | > 5 min | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fbentoml\u002FBentoML?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fbentoml\u002FBentoML) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fbentoml?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fbentoml) | [Gallery](https:\u002F\u002Fwww.bentoml.com\u002Fgallery) | 🟡 | 👍 | 🆓 $10 |\n\nIt goes without saying that these platforms can usually do more than LLM serving**\n\n### 🧮 Serverless Compute Pricing & Limits – Lambda vs Modal (on CPU)\n\n| Platform | 💵 Compute Unit | 📥 Per-Request Fee | 🆓 Free Tier | ⏱️ Max Timeout | 🚦 Concurrency Limit |\n|------------------------|----------------------------------------------------|----------------------------------------------------|------------------------------------------------------|------------------------------------------|-------------------------------------------------------------------|\n| **AWS Lambda + API GW**| GB-sec @ $0.000016667 | $0.20\u002FM Lambda + $1.00\u002FM HTTP API calls | 1M req + 400k GB-s\u002Fmo (12 mo) + 1M API calls\u002Fmo | 15 min | 1,000 per region (can request more) |\n| **Modal** | CPU-s @ $0.0000131 + GiB-s @ $0.00000222 | ❌ No per-request fee | $30\u002Fmo compute credits (Starter) | Func: 24h ⎮ HTTP: 150s → 303 redirect | Starter: 100 containers \u002F 200 req\u002Fs ⎮ Team: 1,000 containers |\n\n\n### Access Off-the-Shelf OS Models (via API):\n\n| Platform\u002FTool | Released | GitHub |\n| --------------------------------------- | -------- | ----------- |\n| [Together.ai](https:\u002F\u002FTogether.ai) | N\u002FA | 🔴 |\n| [Fireworks.ai](https:\u002F\u002FFireworks.ai) | N\u002FA | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Ffw-ai?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Ffw-ai) |\n| [Replicate](https:\u002F\u002Freplicate.com\u002F) | 2019 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Freplicate?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Freplicate) |\n| [Groq](https:\u002F\u002Fgroq.com\u002F) | N\u002FA | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fgroq?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fgroq) |\n| [DeepInfra](https:\u002F\u002Fdeepinfra.com\u002F) | N\u002FA | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fdeepinfra?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fdeepinfra) |\n| [Bedrock](https:\u002F\u002Faws.amazon.com\u002Fbedrock) | N\u002FA | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Faws-samples?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Faws-samples\u002Famazon-bedrock-workshop) |\n| [Lepton](https:\u002F\u002Fwww.lepton.ai\u002F) | N\u002FA | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fleptonai?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fleptonai) |\n| [Fal.ai](https:\u002F\u002Ffal.ai\u002F) | N\u002FA | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Ffal-ai?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Ffal-ai) |\n| [VertexAI](https:\u002F\u002Fcloud.google.com\u002Fvertex-ai) | N\u002FA | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FGoogleCloudPlatform?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FGoogleCloudPlatform\u002Fvertex-ai-samples) |\n\n### Local Inference \n\n| Framework | Browser Chat 🖥️ | Organization | Open Source | GitHub |\n|-----------------------------------------------|------------------|-------------------|-------------|--------|\n| [Llama.cpp](https:\u002F\u002Fgithub.com\u002Fggerganov\u002Fllama.cpp) | ❌ | ggerganov | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fggerganov\u002Fllama.cpp?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fggerganov\u002Fllama.cpp) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fggerganov?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fggerganov) |\n| [Ollama](https:\u002F\u002Follama.com\u002F) | ❌ | Ollama | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Follama\u002Follama?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Follama\u002Follama) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Follama?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Follama) |\n| [gpt4all](https:\u002F\u002Fwww.nomic.ai\u002Fgpt4all) | ✅ | Nomic.ai | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fnomic-ai\u002Fgpt4all?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fnomic-ai\u002Fgpt4all) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fnomic-ai?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fnomic-ai) |\n| [LMStudio](https:\u002F\u002Flmstudio.ai\u002F) | ✅ | LMStudio AI | 🔴 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Flmstudio-ai?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Flmstudio-ai) |\n| [OpenLLM](https:\u002F\u002Fgithub.com\u002Fbentoml\u002FOpenLLM) | ✅ | BentoML | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fbentoml\u002FOpenLLM?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fbentoml\u002FOpenLLM) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fbentoml?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fbentoml) |\n\n### LLM Serving Frameworks\n\n| Framework | Open Source | GitHub |\n|----------------------------------------------------------------|-----------------------------------------------------------------------------------------------------------------------------|------------------------------------------------------------------------------------------|\n| [vLLM](https:\u002F\u002Fgithub.com\u002Fvllm-project\u002Fvllm) | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fvllm-project\u002Fvllm?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fvllm-project\u002Fvllm) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fvllm-project?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fvllm-project) |\n| [OpenLLM](https:\u002F\u002Fgithub.com\u002Fbentoml\u002FOpenLLM) | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fbentoml\u002FOpenLLM?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fbentoml\u002FOpenLLM) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fbentoml?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fbentoml) |\n| [TGI (Text Generation Inference)](https:\u002F\u002Fgithub.com\u002Fhuggingface\u002Ftext-generation-inference) | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fhuggingface\u002Ftext-generation-inference?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fhuggingface\u002Ftext-generation-inference) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fhuggingface?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fhuggingface) |\n| [TensorRT LLM](https:\u002F\u002Fdocs.nvidia.com\u002Ftensorrt-llm\u002Findex.html) | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FNVIDIA\u002FTensorRT-LLM?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FNVIDIA\u002FTensorRT-LLM) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FNVIDIA?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FNVIDIA) |\n| [Ray Serve](https:\u002F\u002Fdocs.ray.io\u002Fen\u002Flatest\u002Fserve\u002Findex.html) | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fray-project\u002Fray?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fray-project\u002Fray) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fray-project?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fray-project) |\n| [LMDeploy](https:\u002F\u002Fgithub.com\u002FInternLM\u002Flmdeploy) | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FInternLM\u002Flmdeploy?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FInternLM\u002Flmdeploy) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FInternLM?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FInternLM) |\n| [Ollama](https:\u002F\u002Fgithub.com\u002Follama\u002Follama) | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Follama\u002Follama?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Follama\u002Follama) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Follama?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Follama) |\n| [MLC-LLM](https:\u002F\u002Fgithub.com\u002Fmlc-ai\u002Fmlc-llm) | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fmlc-ai\u002Fmlc-llm?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fmlc-ai\u002Fmlc-llm) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fmlc-ai?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fmlc-ai) |\n\n\n### Building Open-Source LLM Web Chat UIs\n\n| Tool | Organization | Description | Open Source | GitHub |\n|-------------------------------------------------|--------------------|-------------------------------------------------------------------------------------------------|-----------------------------------------------------------------------------------------------------------------------------|------------------------------------------------------------------------------------------|\n| [Text Generation WebUI](https:\u002F\u002Fgithub.com\u002Foobabooga\u002Ftext-generation-webui) | oobabooga | A Gradio web UI for Large Language Models. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Foobabooga\u002Ftext-generation-webui?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Foobabooga\u002Ftext-generation-webui) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Foobabooga?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Foobabooga) |\n| [Jan AI](https:\u002F\u002Fjan.ai\u002F) | Jan HQ | An open source alternative to ChatGPT that runs 100% offline on your computer. Multiple engine support (llama.cpp, TensorRT-LLM). | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fjanhq\u002Fjan?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fjanhq\u002Fjan) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fjanhq?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fjanhq) |\n| [AnythingLLM](https:\u002F\u002Fgithub.com\u002FMintplex-Labs\u002Fanything-llm) | Mintplex Labs | The all-in-one Desktop & Docker AI application with built-in RAG, AI agents, and more. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FMintplex-Labs\u002Fanything-llm?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FMintplex-Labs\u002Fanything-llm) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FMintplex-Labs?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FMintplex-Labs) |\n| [Superagent](https:\u002F\u002Fgithub.com\u002Fsuperagent-ai\u002Fsuperagent) | Superagent AI | Allows developers to add powerful AI assistants to their applications using LLMs and RAG. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fsuperagent-ai\u002Fsuperagent?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fsuperagent-ai\u002Fsuperagent) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fsuperagent-ai?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fsuperagent-ai) |\n| [Bionic-GPT](https:\u002F\u002Fbionic-gpt.com\u002F) | Bionic GPT | A ChatGPT replacement offering generative AI advantages while maintaining strict data confidentiality. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fbionic-gpt\u002Fbionic-gpt?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fbionic-gpt\u002Fbionic-gpt) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fbionic-gpt?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fbionic-gpt) |\n| [Open WebUI](https:\u002F\u002Fgithub.com\u002Fopen-webui\u002Fopen-webui) | Open WebUI | A user-friendly web interface for interacting with Large Language Models (LLMs). | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fopen-webui\u002Fopen-webui?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fopen-webui\u002Fopen-webui) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fopen-webui?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fopen-webui) |\n| [Xyne](https:\u002F\u002Fgithub.com\u002Fxynehq\u002Fxyne) | xynehq | A sleek, minimal web chat interface for interacting with Large Language Models. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fxynehq\u002Fxyne?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fxynehq\u002Fxyne) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fxynehq?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fxynehq) |\n| [Assistant UI](https:\u002F\u002Fgithub.com\u002Fassistant-ui\u002Fassistant-ui) | assistant-ui | An open-source ChatGPT-like interface with a clean and responsive design. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fassistant-ui\u002Fassistant-ui?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fassistant-ui\u002Fassistant-ui) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fassistant-ui?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fassistant-ui) |\n| [Scira](https:\u002F\u002Fgithub.com\u002Fzaidmukaddam\u002Fscira) | zaidmukaddam | An AI-powered search interface that leverages LLMs for intelligent search results. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fzaidmukaddam\u002Fscira?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fzaidmukaddam\u002Fscira) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fzaidmukaddam?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fzaidmukaddam) |\n| [Onyx](https:\u002F\u002Fgithub.com\u002Fonyx-dot-app\u002Fonyx) | onyx-dot-app | A customizable and extendable web chat UI for interacting with large language models. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fonyx-dot-app\u002Fonyx?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fonyx-dot-app\u002Fonyx) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fonyx-dot-app?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fonyx-dot-app) |\n| [NextChat](https:\u002F\u002Fgithub.com\u002FChatGPTNextWeb\u002FNextChat) | ChatGPTNextWeb | A Next.js-based, open-source ChatGPT clone for seamless web interaction. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FChatGPTNextWeb\u002FNextChat?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FChatGPTNextWeb\u002FNextChat) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FChatGPTNextWeb?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FChatGPTNextWeb) |\n\n### Rent GPUs (Fine-Tuning, Deploying, Training)\n\n| Platform | Templates | Beginner Friendly | GitHub |\n|-----------------------------------------------|--------------------------|-------------------|--------|\n| [Brev.dev](https:\u002F\u002Fwww.brev.dev\u002F) | Fine-tuning | ❌ | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fbrevdev?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fbrevdev) |\n| [Modal](https:\u002F\u002Fmodal.com\u002F) | Fine-tuning | ❌ | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fmodal-labs?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fmodal-labs) |\n| [Hyperbolic AI](https:\u002F\u002Fhyperbolic.xyz\u002F) | None | ❌ | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FHyperbolicLabs?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FHyperbolicLabs) |\n| [RunPod](https:\u002F\u002Fwww.runpod.io\u002F) | None | ❌ | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Frunpod?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Frunpod) |\n| [Paperspace](https:\u002F\u002Fwww.paperspace.com\u002F) | Fine-tuning | ✅ | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FPaperspace?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FPaperspace) |\n| [Colab](https:\u002F\u002Fcolab.research.google.com\u002F) | Small models only | ✅ | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fgooglecolab?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fgooglecolab) |\n\n\n### Fine-Tuning with No-Code UI\n\n| Tool | Beginner Friendly | Open Source | GitHub |\n|------------------------------------------------|-------------------|-------------|--------|\n| [Together.ai](https:\u002F\u002Fwww.together.ai\u002Fproducts#fine-tuning) | ✅ | ❌ | N\u002FA |\n| [Hugging Face AutoTrain](https:\u002F\u002Fhuggingface.co\u002Fautotrain) | ✅ | ❌ | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fhuggingface\u002Fautotrain-advanced?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fhuggingface\u002Fautotrain-advanced) |\n| [AutoML](https:\u002F\u002Fgithub.com\u002Fmljar\u002Fautoml-app) | ❌ | ✅ | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fmljar\u002Fautoml-app?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fmljar\u002Fautoml-app) |\n| [LLaMA-Factory](https:\u002F\u002Fgithub.com\u002Fhiyouga\u002FLLaMA-Factory) | ❌ | ✅ | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fhiyouga\u002FLLaMA-Factory?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fhiyouga\u002FLLaMA-Factory) |\n| [H2O LLM Studio](https:\u002F\u002Fgithub.com\u002Fh2oai\u002Fh2o-llmstudio) | ✅ | ✅ | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fh2oai\u002Fh2o-llmstudio?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fh2oai\u002Fh2o-llmstudio) |\n\n### Fine-Tuning Frameworks\n\n| Framework | Open Source | GitHub |\n|-------------------------------------------|-------------|--------|\n| [Axolotl](https:\u002F\u002Faxolotl.ai\u002F) | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Faxolotl-ai-cloud\u002Faxolotl?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Faxolotl-ai-cloud\u002Faxolotl) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Faxolotl-ai-cloud?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Faxolotl-ai-cloud) |\n| [Unsloth](https:\u002F\u002Funsloth.ai\u002F) | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Funslothai\u002Funsloth?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Funslothai\u002Funsloth) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Funslothai?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Funslothai) |\n\n\n### OS Agentic\u002FAI Workflow\n\n| Framework | Open Source | Beginner Friendly | Released | GitHub |\n|--------------------------------------------------|-----------------------------------------------------------------------------------------------------------------------------|-------------------|----------|------------------------------------------------------------------------------------------|\n| [LangChain](https:\u002F\u002Fwww.langchain.com\u002F) | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Flangchain-ai\u002Flangchain?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Flangchain-ai\u002Flangchain) | ✅ | 2022 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Flangchain-ai?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Flangchain-ai) |\n| [LlamaIndex](https:\u002F\u002Fdocs.llamaindex.ai\u002Fen\u002Fstable\u002Fuse_cases\u002Fagents\u002F) | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Frun-llama\u002Fllama_index?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Frun-llama\u002Fllama_index) | ❌ | 2023 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Frun-llama?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Frun-llama) |\n| [Swarms](https:\u002F\u002Fswarms.world\u002F) | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fkyegomez\u002Fswarms?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fkyegomez\u002Fswarms) | ❌ | 2023 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fkyegomez?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fkyegomez) |\n| [CrewAI](https:\u002F\u002Fwww.crewai.com\u002F) | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fcrewaiinc\u002Fcrewai?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fcrewaiinc\u002Fcrewai) | ✅ | 2023 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fcrewaiinc?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fcrewaiinc) |\n| [Autogen](https:\u002F\u002Fmicrosoft.github.io\u002Fautogen\u002F0.2\u002F) | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fmicrosoft\u002Fautogen?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fautogen) | ✅ | 2023 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fmicrosoft?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fmicrosoft) |\n| [AutoChain](https:\u002F\u002Fautochain.forethought.ai\u002F) | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FForethought-Technologies\u002FAutoChain?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FForethought-Technologies\u002FAutoChain) | ❌ | 2023 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FForethought-Technologies?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FForethought-Technologies) |\n| [SuperAGI](https:\u002F\u002Fsuperagi.com\u002F) | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FTransformerOptimus\u002FSuperAGI?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FTransformerOptimus\u002FSuperAGI) | ❌ | 2023 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FTransformerOptimus?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FTransformerOptimus) |\n| [AILegion](https:\u002F\u002Fgithub.com\u002Feumemic\u002Fai-legion) | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Feumemic\u002Fai-legion?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Feumemic\u002Fai-legion) | ❌ | 2023 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Feumemic?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Feumemic) |\n| [MemGPT (Letta)](https:\u002F\u002Fwww.letta.com\u002F) | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fcpacker\u002FMemGPT?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fcpacker\u002FMemGPT) | ❌ | 2023 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fcpacker?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fcpacker) |\n| [uAgents](https:\u002F\u002Fpypi.org\u002Fproject\u002Fuagents\u002F) | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Ffetchai\u002FuAgents?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Ffetchai\u002FuAgents) | ❌ | 2023 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Ffetchai?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Ffetchai) |\n| [AGiXT](https:\u002F\u002Fgithub.com\u002FJosh-XT\u002FAGiXT) | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FJosh-XT\u002FAGiXT?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FJosh-XT\u002FAGiXT) | ❌ | 2023 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FJosh-XT?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FJosh-XT) |\n| [Dify](https:\u002F\u002Fdify.ai\u002F) | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Flanggenius\u002Fdify?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Flanggenius\u002Fdify) | ✅ | 2024 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Flanggenius?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Flanggenius) |\n| [TaskingAI](https:\u002F\u002Fwww.tasking.ai\u002F) | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FTaskingAI\u002FTaskingAI?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FTaskingAI\u002FTaskingAI) | ✅ | 2024 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FTaskingAI?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FTaskingAI) |\n| [Bee Agent Framework](https:\u002F\u002Fi-am-bee.github.io\u002Fbee-agent-framework\u002F#\u002F) | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fi-am-bee\u002Fbee-agent-framework?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fi-am-bee\u002Fbee-agent-framework) | ❌ | 2024 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fi-am-bee?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fi-am-bee) |\n| [Swarms](https:\u002F\u002Fswarms.world\u002F) | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fkyegomez\u002Fswarms?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fkyegomez\u002Fswarms) | ❌ | 2024 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fkyegomez?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fkyegomez) |\n| [IoA](https:\u002F\u002Fopenbmb.github.io\u002FIoA\u002F) | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FOpenBMB\u002FIoA?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FOpenBMB\u002FIoA) | ❌ | 2024 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FOpenBMB?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FOpenBMB) |\n| [Upsonic](https:\u002F\u002Fgithub.com\u002FUpsonic\u002FUpsonic) | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FUpsonic\u002FUpsonic?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FUpsonic\u002FUpsonic) | ❌ | 2024 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FUpsonic?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FUpsonic) |\n| [Parlant](https:\u002F\u002Fgithub.com\u002Femcie-co\u002Fparlant) | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Femcie-co\u002Fparlant?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Femcie-co\u002Fparlant) | ❌ | 2024 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Femcie-co?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Femcie-co) |\n| [Rig](https:\u002F\u002Fgithub.com\u002F0xPlaygrounds\u002Frig) | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002F0xPlaygrounds\u002Frig?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002F0xPlaygrounds\u002Frig) | ❌ | 2024 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002F0xPlaygrounds?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002F0xPlaygrounds) |\n| [eliza](https:\u002F\u002Fgithub.com\u002FelizaOS\u002Feliza) | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FelizaOS\u002Feliza?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FelizaOS\u002Feliza) | ✅ | 2024 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FelizaOS?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FelizaOS) |\n| [TensorZero](https:\u002F\u002Fwww.tensorzero.com\u002F) | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Ftensorzero\u002Ftensorzero?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Ftensorzero\u002Ftensorzero) | ❌ | 2024 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Ftensorzero?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Ftensorzero\u002Ftensorzero) |\n| [AgentDock](https:\u002F\u002Fagentdock.ai\u002F) | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FAgentDock\u002FAgentDock?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FAgentDock\u002FAgentDock) | ✅ | 2025 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FAgentDock?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FAgentDock) |\n\n### Top Agentic Frameworks\n\n| Framework | Open Source | Beginner Friendly | Released | GitHub |\n|-----------------------------------------------------------|-----------------------------------------------------------------------------------------------------------------------------|-------------------|----------|------------------------------------------------------------------------------------------|\n| [LangGraph](https:\u002F\u002Fwww.langchain.com\u002Flanggraph) | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Flangchain-ai\u002Flanggraph?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Flangchain-ai\u002Flanggraph) | ❌ | 2023 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Flangchain-ai?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Flangchain-ai) |\n| [Flowise](https:\u002F\u002Fflowiseai.com\u002F) | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FFlowiseAI\u002FFlowise?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FFlowiseAI\u002FFlowise) | ✅ | 2023 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FFlowiseAI?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FFlowiseAI) |\n| [Langroid](https:\u002F\u002Flangroid.github.io\u002Flangroid\u002F) | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Flangroid\u002Flangroid?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Flangroid\u002Flangroid) | ❌ | 2023 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Flangroid?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Flangroid) |\n| [smolagents](https:\u002F\u002Fgithub.com\u002Fhuggingface\u002Fsmolagents) | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fhuggingface\u002Fsmolagents?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fhuggingface\u002Fsmolagents) | ❌ | 2024 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fhuggingface?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fhuggingface) |\n| [Semantic Kernel](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fsemantic-kernel) | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fmicrosoft\u002Fsemantic-kernel?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fsemantic-kernel) | ❌ | 2023 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fmicrosoft?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fmicrosoft) |\n| [Atomic Agents](https:\u002F\u002Fgithub.com\u002FBrainBlend-AI\u002Fatomic-agents) | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FBrainBlend-AI\u002Fatomic-agents?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FBrainBlend-AI\u002Fatomic-agents) | ❌ | 2024 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FBrainBlend-AI?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FBrainBlend-AI) |\n| [Agno](https:\u002F\u002Fgithub.com\u002Fagno-agi\u002Fagno) | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fpydantic\u002Fpydantic-ai?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fagno-agi\u002Fagno) | ✅ | 2024 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fpydantic?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fagno-agi) |\n| [PydanticAI](https:\u002F\u002Fgithub.com\u002Fpydantic\u002Fpydantic-ai) | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fpydantic\u002Fpydantic-ai?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fpydantic\u002Fpydantic-ai) | ❌ | 2024 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fpydantic?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fpydantic) |\n| [Mastra](https:\u002F\u002Fgithub.com\u002Fmastra-ai\u002Fmastra) | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fmastra-ai\u002Fmastra?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fmastra-ai\u002Fmastra) | ✅ | 2025 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fmastra-ai?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fmastra-ai) |\n\n### Agentic Frameworks: Core Capabilities\n| Framework | Memory & RAG | Multimodality | Multi-agent Support | Observability |\n|----------------|-----------------------------------------------|-----------------------------------------------|-----------------------------------------------|-------------------------------------------------|\n| **AgentDock** | Built-in RAG system; knowledge base integration | 🟢 Multi-modal (text, voice, tools, APIs) | Visual workflow orchestration & agent chains | Comprehensive LLM traceability & credit tracking |\n| **Agno** | Integrated memory & vector DB\u002FRAG | 🟢 Native (text, image, audio, video) | Supervisor-worker roles | Built-in cloud dashboard\u002Flogging |\n| **LangGraph** | Persistent state; easy external integration | 🔸 Primarily text; extendable via nodes | Hierarchical orchestration | LangSmith integration & graph editor |\n| **SmolAgents** | Built-in short-term; custom long-term | 🔸 Vision agents via VLMs | Modular multi-agent composition | Minimal; relies on external logging |\n| **Mastra** | Persistent workflows; native RAG pipelines | 🟢 Multi-modal via integrations | Native multi-agent workflows | Built-in OpenTelemetry dashboards |\n| **Pydantic AI**| DI-based memory & RAG integration | 🔸 Text-first; multimodal via custom DI | Type-safe manual orchestration | Limited; Python logging\u002FOpenTelemetry |\n| **Atomic Agents**| Per-agent memory & RAG (vector DB) | 🟢 Native multi-modal | Explicit chaining of workflows | Minimal; external instrumentation recommended |\n| **Autogen** | Short-term built-in; external long-term | 🔸 Text-mainly; extensible | Emergent, free-form collaboration | Moderate; internal logging, no dashboard |\n| **CrewAI** | Stateful memory & team-based RAG | 🟢 Diverse modalities (text, image, etc.) | Supervisor-led multi-team workflows | Integrated dashboards for logging & monitoring |\n\nPlease see this [google sheet](https:\u002F\u002Fdocs.google.com\u002Fspreadsheets\u002Fd\u002F1zjcww1w0vARZz9Z6GDxNMp-PKyg7iRyNYAnDo59HjzI\u002Fedit?usp=sharing) with more columns. \n\n### Visual AI Agent Builders\n\n| Tool | Organization | Description | Open Source | GitHub |\n|------|--------------|-------------|-------------|--------|\n| [Rivet](https:\u002F\u002Fgithub.com\u002FIronclad\u002Frivet) | Ironclad | A visual builder to design and deploy AI agent workflows. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FIronclad\u002Frivet?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FIronclad\u002Frivet) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FIronclad?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FIronclad) |\n| [PySpur](https:\u002F\u002Fgithub.com\u002FPySpur-Dev\u002Fpyspur) | PySpur-Dev | A tool to build and visualize AI agents seamlessly. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FPySpur-Dev\u002Fpyspur?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FPySpur-Dev\u002Fpyspur) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FPySpur-Dev?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FPySpur-Dev) |\n| [Flowise](https:\u002F\u002Fgithub.com\u002FFlowiseAI\u002FFlowise) | FlowiseAI | A no‑code, visual platform for designing AI agent workflows. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FFlowiseAI\u002FFlowise?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FFlowiseAI\u002FFlowise) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FFlowiseAI?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FFlowiseAI) |\n\n### 💬 Model Call Pricing for agent systems (Text-only (2000 tokens in, 100 token out), Flat Rate)\n\n| Model | 💵 $ \u002F Call | 💯 100 Calls | 🧮 1,000 Calls | 🔁 30,000 Calls |\n|---------------------|------------:|-------------:|---------------:|----------------:|\n| **Gemini Flash 2.0**| $0.00000 | $0.02 | $0.24 | $7.20 |\n| **GPT-4o mini** | $0.00144 | $0.14 | $1.44 | $43.20 |\n| **GPT-4.1** | $0.00480 | $0.48 | $4.80 | $144.00 |\n| **Gemini Pro 2.5** | $0.00350 | $0.35 | $3.50 | $105.00 |\n| **Claude Haiku 3.5**| $0.00200 | $0.20 | $2.00 | $60.00 |\n| **Claude Sonnet 4** | $0.00750 | $0.75 | $7.50 | $225.00 |\n| **GPT-4o** | $0.01200 | $1.20 | $12.00 | $360.00 |\n| **OpenAI o3** | $0.02400 | $2.40 | $24.00 | $720.00 |\n\n### Agentic Tools (for “building”)\n\n| Tool | Organization | Description | Open Source | GitHub |\n|------|--------------|-------------|-------------|--------|\n| [browser-use](https:\u002F\u002Fgithub.com\u002Fbrowser-use\u002Fbrowser-use) | browser-use | Integrates browser functionalities into agentic workflows. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fbrowser-use\u002Fbrowser-use?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fbrowser-use\u002Fbrowser-use) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fbrowser-use?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fbrowser-use) |\n| [code2prompt](https:\u002F\u002Fgithub.com\u002Fmufeedvh\u002Fcode2prompt) | mufeedvh | Converts code snippets into actionable prompts for development. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fmufeedvh\u002Fcode2prompt?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fmufeedvh\u002Fcode2prompt) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fmufeedvh?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fmufeedvh) |\n| [note-gen](https:\u002F\u002Fgithub.com\u002Fcodexu\u002Fnote-gen) | codexu | Automatically generates notes and documentation from your code. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fcodexu\u002Fnote-gen?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fcodexu\u002Fnote-gen) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fcodexu?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fcodexu) |\n| [refly](https:\u002F\u002Fgithub.com\u002Frefly-ai\u002Frefly) | refly-ai | Automates code refactoring and prompt generation tasks. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Frefly-ai\u002Frefly?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Frefly-ai\u002Frefly) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Frefly-ai?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Frefly-ai) |\n| [potpie](https:\u002F\u002Fgithub.com\u002Fpotpie-ai\u002Fpotpie) | potpie-ai | A toolkit for prototyping and building AI agent pipelines. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fpotpie-ai\u002Fpotpie?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fpotpie-ai\u002Fpotpie) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fpotpie-ai?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fpotpie-ai) |\n| [AgentStack](https:\u002F\u002Fgithub.com\u002FAgentOps-AI\u002FAgentStack) | AgentOps-AI | A comprehensive stack for constructing and deploying AI agents. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FAgentOps-AI\u002FAgentStack?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FAgentOps-AI\u002FAgentStack) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FAgentOps-AI?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FAgentOps-AI) |\n| [browser](https:\u002F\u002Fgithub.com\u002Flightpanda-io\u002Fbrowser) | lightpanda-io | A browser‑based tool designed for integrating agentic functionalities. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Flightpanda-io\u002Fbrowser?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Flightpanda-io\u002Fbrowser) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Flightpanda-io?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Flightpanda-io) |\n| [Memary](https:\u002F\u002Fgithub.com\u002Fkingjulio8238\u002FMemary) | kingjulio8238 | A memory module for retaining context in agent workflows. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fkingjulio8238\u002FMemary?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fkingjulio8238\u002FMemary) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fkingjulio8238?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fkingjulio8238) |\n| [open-canvas](https:\u002F\u002Fgithub.com\u002Flangchain-ai\u002Fopen-canvas) | langchain-ai | A visual interface for designing agent workflows with LangChain. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Flangchain-ai\u002Fopen-canvas?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Flangchain-ai\u002Fopen-canvas) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Flangchain-ai?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Flangchain-ai) |\n| [agent-service-toolkit](https:\u002F\u002Fgithub.com\u002FJoshuaC215\u002Fagent-service-toolkit) | JoshuaC215 | A toolkit for building and deploying agent-based services. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FJoshuaC215\u002Fagent-service-toolkit?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FJoshuaC215\u002Fagent-service-toolkit) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FJoshuaC215?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FJoshuaC215) |\n\n\n### Virtual Brains\n\n| Tool | Organization | Description | Open Source | GitHub |\n|------|--------------|-------------|-------------|--------|\n| [Leon](https:\u002F\u002Fgithub.com\u002Fleon-ai\u002Fleon) | leon-ai | An open‑source personal assistant and automation platform powered by AI. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fleon-ai\u002Fleon?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fleon-ai\u002Fleon) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fleon-ai?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fleon-ai) |\n| [Khoj](https:\u002F\u002Fgithub.com\u002Fkhoj-ai\u002Fkhoj) | khoj-ai | A virtual brain for organizing and retrieving your knowledge using AI. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fkhoj-ai\u002Fkhoj?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fkhoj-ai\u002Fkhoj) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fkhoj-ai?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fkhoj-ai) |\n\n\n### AI Agents\n\n| Framework | Organization | Open Source | Released | GitHub |\n|-----------------------------------------------------|------------------------------|-----------------------------------------------------------------------------------------------------------------------------|----------|------------------------------------------------------------------------------------------|\n| [GPT Engineer](https:\u002F\u002Fgptengineer.app\u002F) | GPT Engineer Org | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fgpt-engineer-org\u002Fgpt-engineer?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fgpt-engineer-org\u002Fgpt-engineer) | 2023 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fgpt-engineer-org?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fgpt-engineer-org) |\n| [XAgent](https:\u002F\u002Fgithub.com\u002FOpenBMB\u002FXAgent) | OpenBMB | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FOpenBMB\u002FXAgent?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FOpenBMB\u002FXAgent) | 2023 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FOpenBMB?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FOpenBMB) |\n| [Bolt.new](https:\u002F\u002Fgithub.com\u002Fstackblitz\u002Fbolt.new) | StackBlitz | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fstackblitz\u002Fbolt.new?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fstackblitz\u002Fbolt.new) | 2023 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fstackblitz?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fstackblitz) |\n| [Goose](https:\u002F\u002Fgithub.com\u002Fblock\u002Fgoose) | Block | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fblock\u002Fgoose?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fblock\u002Fgoose) | 2023 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fblock?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fblock) |\n| [AI Hedge Fund](https:\u002F\u002Fgithub.com\u002Fvirattt\u002Fai-hedge-fund) | virattt | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fvirattt\u002Fai-hedge-fund?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fvirattt\u002Fai-hedge-fund) | 2023 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fvirattt?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fvirattt) |\n| [FinRobot](https:\u002F\u002Fai4finance.org\u002F) | AI4Finance Foundation | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FAI4Finance-Foundation\u002FFinRobot?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FAI4Finance-Foundation\u002FFinRobot) | 2024 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FAI4Finance-Foundation?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FAI4Finance-Foundation) |\n| [STORM](https:\u002F\u002Fstorm.genie.stanford.edu\u002F) | Stanford OVAL | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fstanford-oval\u002Fstorm?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fstanford-oval\u002Fstorm) | 2024 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fstanford-oval?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fstanford-oval) |\n| [Multion](https:\u002F\u002Fwww.multion.ai\u002F) | MULTI-ON | 🔴 | N\u002FA | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FMULTI-ON?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FMULTI-ON) |\n| [Minion](https:\u002F\u002Fminion.ai\u002F) | Minion AI | 🔴 | N\u002FA | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fminionai?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fminionai) |\n\n### Long-Term Memory\n\n| Provider | Community | Founded | GitHub | ⭐ Stars | Open Source |\n|--------------|------------------------|-------------|--------------------------------------------------------------------------------------------------|---------|-----------------------------| \n| Letta | 💬 Active dev community| Oct 2023 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fletta-ai?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fletta-ai\u002Fletta) | 17k | ✅ Apache-2.0 |\n| Zep | 🤝 Moderate community | Aug 2024 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fgetzep?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fgetzep\u002Fgraphiti) | 11.6k | ⚠️ Graphiti CE (Apache-2.0) |\n| MemoRAG | 🧪 Small research group| Sep 2024 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fqhjqhj00?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fqhjqhj00\u002FMemoRAG) | 1.8k | ✅ Apache-2.0 |\n| Memary | 🧠 Niche community | April 2024 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fkingjulio8238?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fkingjulio8238\u002FMemary) | 2.3k | ✅ MIT |\n| Cognee | 🔄 Moderate | Aug 2023 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Ftopoteretes?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Ftopoteretes\u002Fcognee) | 5.8k | ✅ Apache-2.0 |\n| Mem0 | 🚀 Fast-growing | June 2023 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fmem0ai?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fmem0ai\u002Fmem0) | 35.2k | ✅ Apache-2.0 |\n\n#### Memory Features Comparison\n\n| Provider | Based | Optional KG | Self-Editing \u002F Agentic | Rolling Summaries | Categories |\n|-----------|---------|-------------|-------------------------|-------------------------------|------------|\n| Letta | 🧮 Vector | ⚠️ Partial | ✅ Yes | ⚠️ Partial (memory blocks) | ✅ Yes |\n| Zep | 🧠 KG | - | ✅ Yes | ✅ Auto chat summarization | ✅ Yes |\n| MemoRAG | 🧮 Vector | ❌ No | ✅ Yes | ❌ Uses long-range model | ❌ No |\n| Memary | 🧠 KG | - | ✅ Yes | ⚠️ Plans “rewind” feature | ✅ Yes |\n| Cognee | 🧠 KG | - | ✅ Yes | ❌ No auto summaries | ✅ Yes |\n| Mem0 | 🧮 Vector | ❌ No | ✅ Yes | ❌ Not explicit | ✅ Yes |\n\n#### Enterprise Security (Cloud-Based Use)\n\n| Provider | Enterprise Security |\n|-----------|--------------------------------------------------------------------------------------------------------|\n| Mem0 | 🔐 Hosted with encryption, org\u002Fproject roles, GDPR-friendly delete. Uses Graphlit (SOC 2 not stated). |\n| Letta | ☁️ Self-hosted or managed server. User auth & ID-partitioned memory. Graphlit-based. No public SSO details. |\n| Zep | ✅ SOC 2 Type 2. Encrypted at rest\u002Fin transit, access controls, JWT, and deletion API (\"Right to be Forgotten\"). |\n| MemoRAG | 🏠 Self-host |\n| Memary | 🏠 Self-host |\n| Cognee | 🏠 Self-host |\n\n\n#### Pricing by Monthly Message Volume\n\n| Provider | 1K msgs\u002Fmo | 10K msgs\u002Fmo | 100K msgs\u002Fmo | 1M msgs\u002Fmo |\n|------------|------------------|---------------------|---------------------------|---------------------------|\n| Mem0 | 🆓 Free | 🆓 Free–$29 | 💵 $249 | 🏢 Enterprise (custom) |\n| Letta | 🆓 Free | 💵 $20 | 💵 $750 | 🏢 Enterprise (custom) |\n| Zep | 🆓 Free | 🆓 Free | 💵 ~ $112.50 | 💵 ~ $1,237 |\n| MemoRAG | 💻 GPU Server (~$150–300\u002Fmo) | 💻 GPU Server (~$150–300\u002Fmo) | 💻 Multi-GPU ($500+) | 🖥️ Cluster ($1K+\u002Fmo) |\n| Self-host | 🖥️ Small VM (~$15\u002Fmo) | 🖥️ Small VM (~$15–20\u002Fmo) | 🖥️ Medium VM ($50–$100\u002Fmo) | 🖥️ Large VM ($200+\u002Fmo) |\n\n### Evaluation Frameworks and add-ons\n\n| Framework | Open Source | Beginner Friendly | Released | GitHub |\n| ------------------------------------------------------------------ | ----------------------------------------------------------------------------------------------------------------------------------------------------------------- | ----------------- | -------- | ------------------------------------------------------------------------------------------------------------------------------------------------------ |\n| [TruLens](https:\u002F\u002Fgithub.com\u002Ftruera\u002Ftrulens) | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Ftruera\u002Ftrulens?style=flat-square\\&color=purple)](https:\u002F\u002Fgithub.com\u002Ftruera\u002Ftrulens) | ✅ | 2023 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Ftruera?style=flat-square\\&color=teal)](https:\u002F\u002Fgithub.com\u002Ftruera) |\n| [Promptfoo](https:\u002F\u002Fgithub.com\u002Fpromptfoo\u002Fpromptfoo) | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fpromptfoo\u002Fpromptfoo?style=flat-square\\&color=purple)](https:\u002F\u002Fgithub.com\u002Fpromptfoo\u002Fpromptfoo) | ✅ | 2023 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fpromptfoo?style=flat-square\\&color=teal)](https:\u002F\u002Fgithub.com\u002Fpromptfoo) |\n| [DeepEval](https:\u002F\u002Fgithub.com\u002Fconfident-ai\u002Fdeepeval) | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fconfident-ai\u002Fdeepeval?style=flat-square\\&color=purple)](https:\u002F\u002Fgithub.com\u002Fconfident-ai\u002Fdeepeval) | ✅ | 2024 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fconfident-ai?style=flat-square\\&color=teal)](https:\u002F\u002Fgithub.com\u002Fconfident-ai) |\n| [RAGAS](https:\u002F\u002Fgithub.com\u002Fexplodinggradients\u002Fragas) | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fexplodinggradients\u002Fragas?style=flat-square\\&color=purple)](https:\u002F\u002Fgithub.com\u002Fexplodinggradients\u002Fragas) | ❌ | 2023 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fexplodinggradients?style=flat-square\\&color=teal)](https:\u002F\u002Fgithub.com\u002Fexplodinggradients) |\n| [OpenAI Evals](https:\u002F\u002Fgithub.com\u002Fopenai\u002Fevals) | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fopenai\u002Fevals?style=flat-square\\&color=purple)](https:\u002F\u002Fgithub.com\u002Fopenai\u002Fevals) | ❌ | 2023 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fopenai?style=flat-square\\&color=teal)](https:\u002F\u002Fgithub.com\u002Fopenai) |\n| [LangChain OpenEvals](https:\u002F\u002Fgithub.com\u002Flangchain-ai\u002Fopenevals) | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Flangchain-ai\u002Fopenevals?style=flat-square\\&color=purple)](https:\u002F\u002Fgithub.com\u002Flangchain-ai\u002Fopenevals) | ✅ | 2025 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Flangchain-ai?style=flat-square\\&color=teal)](https:\u002F\u002Fgithub.com\u002Flangchain-ai) |\n| [LangChain AgentEvals](https:\u002F\u002Fgithub.com\u002Flangchain-ai\u002Fagentevals) | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Flangchain-ai\u002Fagentevals?style=flat-square\\&color=purple)](https:\u002F\u002Fgithub.com\u002Flangchain-ai\u002Fagentevals) | ❌ | 2025 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Flangchain-ai?style=flat-square\\&color=teal)](https:\u002F\u002Fgithub.com\u002Flangchain-ai) |\n| [LlamaIndex Eval](https:\u002F\u002Fgithub.com\u002Frun-llama\u002Fllama_index) | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Frun-llama\u002Fllama_index?style=flat-square\\&color=purple)](https:\u002F\u002Fgithub.com\u002Frun-llama\u002Fllama_index) | ✅ | 2023 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Frun-llama?style=flat-square\\&color=teal)](https:\u002F\u002Fgithub.com\u002Frun-llama) |\n\n### Evaluation Frameworks: Core Differences\n\n| Framework | Pytest \u002F CLI Runner | Metrics Ready-made | Synthetic Data Gen | Offline Judge | Model-Agnostic | Safety | Red-Team | **Custom Metrics (setup speed)** |\n| ------------------- | --------------------- | ------------------ | ---------------------------- | ------------- | -------------- | ------ | -------- | -------------------------------------------------- |\n| **DeepEval** | 🟢 `deepeval test` | **40 +** | 🟢 `deepeval create-dataset` | 🟢 | 🟢 | 🟢 | 🟢 | 🟢 **G-Eval builder — minutes (one function)** |\n| **RAGAS** | ✖ (script asserts) | 6 core RAG + 🔸 | 🟢 KG-based Q-gen | 🟢 | 🟢 | 🔸 DIY | ✖ | 🟢 **`AspectCritic` one-liner — minutes** |\n| **MLflow Evaluate** | ✖ (`mlflow.evaluate`) | 3-4 | ✖ BYO | 🔸 possible | 🔸 | 🟢 | ✖ | 🟢 **Subclass scorer — few lines, \\~hour** |\n| **OpenAI Evals** | 🟢 CLI orchestrator | \\~10 templates | 🔸 helper script | ✖ | 🟢 | ✖ | ✖ | 🟢 **Full Python\u002FYAML eval — flexible but slower** |\n\n\n### Co-Pilots\n\n| Framework | Open Source | GitHub |\n|--------------------------------------------|-----------------------------------------------------------------------------------------------------------------------------|------------------------------------------------------------------------------------------|\n| [Aider](https:\u002F\u002Faider.chat\u002F) | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FAider-AI\u002Faider?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FAider-AI\u002Faider) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FAider-AI?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FAider-AI) |\n| [Cursor](https:\u002F\u002Fwww.cursor.com\u002F) | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fgetcursor\u002Fcursor?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fgetcursor\u002Fcursor) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fgetcursor?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fgetcursor) |\n| [Continue](https:\u002F\u002Fdocs.continue.dev\u002F) | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fcontinuedev\u002Fcontinue?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fcontinuedev\u002Fcontinue) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fcontinuedev?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fcontinuedev) |\n\n\n### Voice API\n\n| Framework | Open Source | GitHub |\n|--------------------------------------------|-----------------------------------------------------------------------------------------------------------------------------|------------------------------------------------------------------------------------------|\n| [VAPI.ai](https:\u002F\u002Fvapi.ai\u002F) | 🔴 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fvapiai?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fvapiai) |\n| [Bland.ai](https:\u002F\u002Fwww.bland.ai\u002F) | 🔴 | N\u002FA |\n| [CallAnnie](https:\u002F\u002Fcallannie.ai\u002F) | 🔴 | N\u002FA |\n| [RealtimeTTS](https:\u002F\u002Fgithub.com\u002FKoljaB\u002FRealtimeTTS) | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FKoljaB\u002FRealtimeTTS?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FKoljaB\u002FRealtimeTTS) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FKoljaB?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FKoljaB) |\n| [RealtimeSTT](https:\u002F\u002Fgithub.com\u002FKoljaB\u002FRealtimeSTT) | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FKoljaB\u002FRealtimeSTT?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FKoljaB\u002FRealtimeSTT) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FKoljaB?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FKoljaB) |\n| [Coqui TTS](https:\u002F\u002Fgithub.com\u002Fcoqui-ai\u002FTTS) | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fcoqui-ai\u002FTTS?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fcoqui-ai\u002FTTS) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fcoqui-ai?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fcoqui-ai) |\n\n### Open Source TTS Models\n\n| Model | License | Stars\u002FLikes | Downloads (Last Month) | Repository |\n|----------------------------------------------|------------------------------|----------------|-------------------------|----------------|\n| [Kokoro-82M](https:\u002F\u002Fhuggingface.co\u002Fhexgrad\u002FKokoro-82M) | Apache 2.0 | ⭐ 3.16k (HF) | 📥 557,392 | [Hugging Face](https:\u002F\u002Fhuggingface.co\u002Fhexgrad\u002FKokoro-82M) |\n| [Zonos-v0.1-transformer](https:\u002F\u002Fhuggingface.co\u002FZyphra\u002FZonos-v0.1-transformer) | Apache 2.0 | ⭐ 249 (HF) | 📥 24,240 | [Hugging Face](https:\u002F\u002Fhuggingface.co\u002FZyphra\u002FZonos-v0.1-transformer) |\n| [XTTS-v2](https:\u002F\u002Fhuggingface.co\u002Fcoqui\u002FXTTS-v2) | Non-Commercial | ❤️ 368 (HF) | 📥 2,545,850 | [Hugging Face](https:\u002F\u002Fhuggingface.co\u002Fcoqui\u002FXTTS-v2) |\n| [ChatTTS](https:\u002F\u002Fgithub.com\u002F2noise\u002FChatTTS) | AGPL-3.0 | N\u002FA | N\u002FA | [GitHub](https:\u002F\u002Fgithub.com\u002F2noise\u002FChatTTS) |\n| [MeloTTS](https:\u002F\u002Fgithub.com\u002Fmyshell-ai\u002FMeloTTS) | MIT | N\u002FA | N\u002FA | [GitHub](https:\u002F\u002Fgithub.com\u002Fmyshell-ai\u002FMeloTTS) |\n\nFor more TTS models and rankings, check out the [TTS Leaderboard](https:\u002F\u002Fhuggingface.co\u002Fspaces\u002FPendrokar\u002FTTS-Spaces-Arena).\n\n### LLM Application Frameworks\n\n| Tool | Organization | Description | Open Source | GitHub |\n|------|--------------|-------------|-------------|--------|\n| [Eino](https:\u002F\u002Fgithub.com\u002Fcloudwego\u002Feino) | CloudWeGo | A lightweight LLM application framework for scalable AI solutions. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fcloudwego\u002Feino?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fcloudwego\u002Feino) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fcloudwego?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fcloudwego) |\n| [Conversation Knowledge Mining Solution Accelerator](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002FConversation-Knowledge-Mining-Solution-Accelerator) | Microsoft | A solution accelerator for integrating conversation intelligence and knowledge mining using LLMs. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fmicrosoft\u002FConversation-Knowledge-Mining-Solution-Accelerator?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002FConversation-Knowledge-Mining-Solution-Accelerator) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fmicrosoft?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fmicrosoft) |\n| [Olmocr](https:\u002F\u002Fgithub.com\u002Fallenai\u002Folmocr) | AllenAI | An OCR framework optimized for integration with language models. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fallenai\u002Folmocr?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fallenai\u002Folmocr) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fallenai?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fallenai) |\n| [PDFMathTranslate](https:\u002F\u002Fgithub.com\u002FByaidu\u002FPDFMathTranslate) | Byaidu | A tool for converting and translating mathematical content in PDFs using LLMs. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FByaidu\u002FPDFMathTranslate?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FByaidu\u002FPDFMathTranslate) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FByaidu?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FByaidu) |\n| [Podcastfy](https:\u002F\u002Fgithub.com\u002Fsouzatharsis\u002Fpodcastfy) | souzatharsis | A tool to generate podcasts from written content using LLMs. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fsouzatharsis\u002Fpodcastfy?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fsouzatharsis\u002Fpodcastfy) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fsouzatharsis?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fsouzatharsis) |\n| [Pandas AI](https:\u002F\u002Fgithub.com\u002Fsinaptik-ai\u002Fpandas-ai) | sinaptik-ai | Brings LLM-powered analytics to pandas dataframes. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fsinaptik-ai\u002Fpandas-ai?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fsinaptik-ai\u002Fpandas-ai) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fsinaptik-ai?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fsinaptik-ai) |\n| [Ramalama](https:\u002F\u002Fgithub.com\u002Fcontainers\u002Framalama) | containers | An LLM application framework for containerized deployment of AI solutions. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fcontainers\u002Framalama?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fcontainers\u002Framalama) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fcontainers?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fcontainers) |\n| [Robyn](https:\u002F\u002Fgithub.com\u002Ffacebookexperimental\u002FRobyn) | facebookexperimental | A scalable framework for building LLM applications from Facebook Experimental. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Ffacebookexperimental\u002FRobyn?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Ffacebookexperimental\u002FRobyn) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Ffacebookexperimental?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Ffacebookexperimental) |\n| [ExtractThinker](https:\u002F\u002Fgithub.com\u002Fenoch3712\u002FExtractThinker) | enoch3712 | A tool for extracting and synthesizing insights from textual data using LLMs. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fenoch3712\u002FExtractThinker?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fenoch3712\u002FExtractThinker) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fenoch3712?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fenoch3712) |\n\n\n### OS RAG Frameworks\n\n| Framework | Organization | Open Source | Released | GitHub |\n|-----------------------------------------------------|--------------|-----------------------------------------------------------------------------------------------------------------------------|----------|------------------------------------------------------------------------------------------|\n| [Haystack](https:\u002F\u002Fhaystack.deepset.ai\u002F) | deepset.ai | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fdeepset-ai\u002Fhaystack?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fdeepset-ai\u002Fhaystack) | 2023 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fdeepset-ai?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fdeepset-ai) |\n| [RAGflow](https:\u002F\u002Fragflow.io\u002F) | Infiniflow | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Finfiniflow\u002Fragflow?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Finfiniflow\u002Fragflow) | 2024 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Finfiniflow?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Finfiniflow) |\n| [txtai](https:\u002F\u002Fneuml.github.io\u002Ftxtai\u002F) | Neuml | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fneuml\u002Ftxtai?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fneuml\u002Ftxtai) | 2022 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fneuml?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fneuml) |\n| [LLM App](https:\u002F\u002Fpathway.com\u002Fdevelopers\u002Ftemplates) | Pathway | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fpathwaycom\u002Fllm-app?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fpathwaycom\u002Fllm-app) | 2023 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fpathwaycom?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fpathwaycom) |\n| [Cognita](https:\u002F\u002Fcognita.truefoundry.com\u002F) | Truefoundry | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Ftruefoundry\u002Fcognita?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Ftruefoundry\u002Fcognita) | 2024 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Ftruefoundry?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Ftruefoundry) |\n| [R2R](https:\u002F\u002Fr2r-docs.sciphi.ai\u002Fintroduction) | SciPhi AI | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FSciPhi-AI\u002FR2R?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FSciPhi-AI\u002FR2R) | 2024 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FSciPhi-AI?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FSciPhi-AI) |\n| [Raptor](https:\u002F\u002Farxiv.org\u002Fabs\u002F2401.18059) | Parth Sarthi | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fparthsarthi03\u002Fraptor?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fparthsarthi03\u002Fraptor) | 2024 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fparthsarthi03?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fparthsarthi03) |\n| [LightRAG](https:\u002F\u002Fgithub.com\u002FHKUDS\u002FLightRAG) | HKUDS | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FHKUDS\u002FLightRAG?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FHKUDS\u002FLightRAG) | 2023 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FHKUDS?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FHKUDS) |\n| [PIKE-RAG](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002FPIKE-RAG) | Microsoft | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fmicrosoft\u002FPIKE-RAG?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002FPIKE-RAG) | 2024 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fmicrosoft?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fmicrosoft) |\n| [KAG](https:\u002F\u002Fgithub.com\u002FOpenSPG\u002FKAG) | OpenSPG | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FOpenSPG\u002FKAG?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FOpenSPG\u002FKAG) | 2024 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FOpenSPG?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FOpenSPG) |\n| [MemoRAG](https:\u002F\u002Fgithub.com\u002Fqhjqhj00\u002FMemoRAG) | qhjqhj00 | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fqhjqhj00\u002FMemoRAG?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fqhjqhj00\u002FMemoRAG) | 2023 | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fqhjqhj00?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fqhjqhj00) |\n\nSee [RAG_Techniques](https:\u002F\u002Fgithub.com\u002FNirDiamant\u002FRAG_Techniques) if you get stuck (not always needed)\n\n### 🔍 Vector DBs – FOSS, Performance, Pricing, DevX\n\n| Vector DB | License | ⚡ Perf \u002F Throughput | ⏱️ Latency (Real-World) | ☁️ Cloud Pricing \u002F Free Tier | 💻 Dev Experience |\n|------------------|----------------|-------------------------------------------------------------------------------------|--------------------------------------------------------------|-------------------------------------------------------------------------------|----------------------------------------------------------------------------------------------------|\n| **Qdrant** | Apache 2.0 | 🥇 Highest RPS, lowest latency in single-node bench (≥4× vs prev. run) | p95 \u003C 10ms for 1M vecs (1 thread) | Always-on 1GB free; pay-go ≈ $0.014\u002Fhr | REST + gRPC; 7 lang clients; filter-aware HNSW; hybrid support; Python “embedded” mode |\n| **Milvus \u002F Zilliz Cloud** | Apache 2.0 | 🚀 Fastest index build; RPS trails Qdrant for high-dim vecs | p95 ≈ 10–20ms @ 1M 768-dim (DiskANN, vendor data) | 5GB free; serverless $0.30\u002FGB-mo; dedicated from $99\u002Fmo | New SDK v2 (async, schema cache); Python\u002FGo\u002FJava\u002FNode support |\n| **Weaviate** | BSD-3-Clause | ⚙️ Least bench gains, but decent recall (95%+) and throughput | “Low-ms” claimed; users report 100–300ms if misconfigured | Starts $25\u002Fmo; 14-day sandbox free | GraphQL + REST; strong SDKs (Py\u002FTS\u002FGo\u002FJava); easy RAG + hybrid templates |\n| **pgvector** | MIT | 🔥 28× lower p95 & 16× higher QPS vs Pinecone s1 @ 99% recall (50M) | p95 \u003C 50ms @ 50M 768-dim (Timescale test) | Neon\u002FSupabase offer free Postgres with pgvector (0.5–1GB, ~200h CPU) | Pure SQL; supports joins + ACID; great for hybrid text + dense queries |\n| **Redis 8 Vector** | AGPLv3 \u002F RSAL \u002F SSPL | 🧵 3.4× higher QPS vs Qdrant, 4× vs Milvus @ ≥0.98 recall | Sub-ms avg, \u003C10ms under load (vendor); 9.7× lower than Aurora+pgvector | Redis Cloud: 30MB free, pay-go from $5\u002Fmo; Flex $0.007\u002Fhr | Redis Vector Library + RAG helpers; OM clients for .NET\u002FPy\u002FJS; fast setup |\n\n### 💾 Vector DB Cloud Pricing (2000-char Chunks, ~768-dim)\n\n| # Chunks | Data Size | 🟣 Milvus \u002F Zilliz Cloud (Serverless) | 🟢 Qdrant Cloud | 🟡 Weaviate Cloud (“Standard”) |\n|----------|-------------|------------------------------------------------------|-------------------------------------------------|---------------------------------------------------------------|\n| **10k** | ~0.07 GB | 🆓 Free – within 5 GB tier | 🆓 Free – fits in 1 GB RAM \u002F 4 GB disk | $25 base + $1.2 dim fee ≈ **$26** |\n| **100k** | ~0.67 GB | 🆓 Still under 5 GB | 🆓 Fits with compression in 4 GB disk | $25 + $12.0 dim fee ≈ **$37** |\n| **1M** | ~6.7 GB | 💵 ≈ $2 storage; add vCU fees or $99 dedicated | 💵 Needs 10 GB cluster → ≈ **$20\u002Fmo** | $25 + $120.5 dim fee ≈ **$145** |\n| **10M** | ~67 GB | 💵 ≈ $20 storage; + compute: **$100–150 total** | 💵 Needs 64+ GB → **$120–150\u002Fmo** estimate | $25 + $1,204 dim fee ≈ **$1,230** |\n\n### 🧠 Embedding Cost – OpenAI (Small Model, per Chunk Size)\n\n| # Chunks | 📏 1,000 Chars | 📏 2,000 Chars | 📏 3,000 Chars |\n|--------------|---------------:|---------------:|---------------:|\n| **1,000** | $0.01 | $0.01 | $0.01 |\n| **10,000** | $0.05 | $0.10 | $0.15 |\n| **100,000** | $0.50 | $1.00 | $1.50 |\n| **1,000,000** | $5.00 | $10.00 | $15.00 |\n| **10,000,000**| $50.00 | $100.00 | $150.00 |\n\n\n### AI Tools (for “using”)\n\n| Tool | Organization | Description | Open Source | GitHub |\n|------|--------------|-------------|-------------|--------|\n| [magic-resume](https:\u002F\u002Fgithub.com\u002FJOYCEQL\u002Fmagic-resume) | JOYCEQL | An AI-powered tool for generating resumes. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FJOYCEQL\u002Fmagic-resume?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FJOYCEQL\u002Fmagic-resume) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FJOYCEQL?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FJOYCEQL) |\n| [VideoCaptioner](https:\u002F\u002Fgithub.com\u002FWEIFENG2333\u002FVideoCaptioner) | WEIFENG2333 | An AI tool for automatically generating video captions. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FWEIFENG2333\u002FVideoCaptioner?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FWEIFENG2333\u002FVideoCaptioner) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FWEIFENG2333?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FWEIFENG2333) |\n| [DeepSeekAI](https:\u002F\u002Fgithub.com\u002FDeepLifeStudio\u002FDeepSeekAI) | DeepLifeStudio | Browser extension for invoking the DeepSeek AI large model. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FDeepLifeStudio\u002FDeepSeekAI?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FDeepLifeStudio\u002FDeepSeekAI) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FDeepLifeStudio?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FDeepLifeStudio) |\n| [logocreator](https:\u002F\u002Fgithub.com\u002FNutlope\u002Flogocreator) | Nutlope | A tool for creating logos using AI. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FNutlope\u002Flogocreator?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FNutlope\u002Flogocreator) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FNutlope?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FNutlope) |\n| [blinkshot](https:\u002F\u002Fgithub.com\u002FNutlope\u002Fblinkshot) | Nutlope | An AI-powered tool for capturing and enhancing screenshots. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FNutlope\u002Fblinkshot?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FNutlope\u002Fblinkshot) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FNutlope?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FNutlope) |\n| [pollinations](https:\u002F\u002Fgithub.com\u002Fpollinations\u002Fpollinations) | pollinations | A tool for generating creative images and artwork using AI. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fpollinations\u002Fpollinations?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fpollinations\u002Fpollinations) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fpollinations?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fpollinations) |\n| [PromptWizard](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002FPromptWizard) | microsoft | A tool to generate, manage, and optimize prompts for AI models. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fmicrosoft\u002FPromptWizard?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002FPromptWizard) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fmicrosoft?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fmicrosoft) |\n| [Open-Interface](https:\u002F\u002Fgithub.com\u002FAmberSahdev\u002FOpen-Interface) | AmberSahdev | Control Any Computer Using LLMs. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FAmberSahdev\u002FOpen-Interface?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FAmberSahdev\u002FOpen-Interface) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FAmberSahdev?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FAmberSahdev) |\n| [wut](https:\u002F\u002Fgithub.com\u002Fshobrook\u002Fwut) | shobrook | LLM for the terminal | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fshobrook\u002Fwut?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fshobrook\u002Fwut) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fshobrook?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fshobrook) |\n\n\n### Training\u002FOptimization\n\n| Tool | Organization | Description | Open Source | GitHub |\n|------|--------------|-------------|-------------|--------|\n| [transformerlab-app](https:\u002F\u002Fgithub.com\u002Ftransformerlab\u002Ftransformerlab-app) | transformerlab | An application for training and optimizing transformer models. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Ftransformerlab\u002Ftransformerlab-app?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Ftransformerlab\u002Ftransformerlab-app) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Ftransformerlab?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Ftransformerlab) |\n| [fluxgym](https:\u002F\u002Fgithub.com\u002Fcocktailpeanut\u002Ffluxgym) | cocktailpeanut | A gym environment for reinforcement learning training and optimization. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fcocktailpeanut\u002Ffluxgym?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fcocktailpeanut\u002Ffluxgym) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fcocktailpeanut?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fcocktailpeanut) |\n| [AutoGPTQ](https:\u002F\u002Fgithub.com\u002FAutoGPTQ\u002FAutoGPTQ) | AutoGPTQ | A tool for automating GPT quantization and optimization. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FAutoGPTQ\u002FAutoGPTQ?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FAutoGPTQ\u002FAutoGPTQ) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FAutoGPTQ?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FAutoGPTQ) |\n\n\n### AI Models\n\n| Tool | Organization | Description | Open Source | GitHub |\n|------|--------------|-------------|-------------|--------|\n| [WALDO](https:\u002F\u002Fgithub.com\u002Fstephansturges\u002FWALDO) | stephansturges | An AI model for visual reasoning and object detection. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fstephansturges\u002FWALDO?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fstephansturges\u002FWALDO) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fstephansturges?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fstephansturges) |\n| [Janus](https:\u002F\u002Fgithub.com\u002Fdeepseek-ai\u002FJanus) | deepseek-ai | A multi-modal AI model for advanced data processing. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fdeepseek-ai\u002FJanus?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fdeepseek-ai\u002FJanus) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fdeepseek-ai?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fdeepseek-ai) |\n| [ModernBERT](https:\u002F\u002Fgithub.com\u002FAnswerDotAI\u002FModernBERT) | AnswerDotAI | A modernized version of BERT for natural language processing tasks. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FAnswerDotAI\u002FModernBERT?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FAnswerDotAI\u002FModernBERT) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FAnswerDotAI?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FAnswerDotAI) |\n| [Magma](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002FMagma) | microsoft | A scalable AI model for large-scale data analysis. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fmicrosoft\u002FMagma?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002FMagma) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fmicrosoft?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fmicrosoft) |\n| [Cosmos-Nemotron](https:\u002F\u002Fgithub.com\u002FNVlabs\u002FCosmos-Nemotron) | NVlabs | An AI model for advanced image and video processing. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FNVlabs\u002FCosmos-Nemotron?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FNVlabs\u002FCosmos-Nemotron) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FNVlabs?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FNVlabs) |\n| [Paints-UNDO](https:\u002F\u002Fgithub.com\u002Flllyasviel\u002FPaints-UNDO) | lllyasviel | An interactive AI model for image generation and editing. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Flllyasviel\u002FPaints-UNDO?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Flllyasviel\u002FPaints-UNDO) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Flllyasviel?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Flllyasviel) |\n\n\n### Monitoring\n\n| Tool | Organization | Description | Open Source | GitHub |\n|------|--------------|-------------|-------------|--------|\n| [helicone](https:\u002F\u002Fgithub.com\u002FHelicone\u002Fhelicone) | Helicone | A platform for monitoring and analyzing AI model performance. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FHelicone\u002Fhelicone?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FHelicone\u002Fhelicone) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FHelicone?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FHelicone) |\n| [langwatch](https:\u002F\u002Fgithub.com\u002Flangwatch\u002Flangwatch) | langwatch | A tool for monitoring outputs and performance of language models. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Flangwatch\u002Flangwatch?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Flangwatch\u002Flangwatch) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Flangwatch?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Flangwatch) |\n\n\n### Infrastructure\n\n| Tool | Organization | Description | Open Source | GitHub |\n|------|--------------|-------------|-------------|--------|\n| [gpustack](https:\u002F\u002Fgithub.com\u002Fgpustack\u002Fgpustack) | gpustack | A toolkit for managing GPU infrastructure for AI workloads. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fgpustack\u002Fgpustack?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fgpustack\u002Fgpustack) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fgpustack?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fgpustack) |\n| [harbor](https:\u002F\u002Fgithub.com\u002Fav\u002Fharbor) | av | A repository for containerized AI infrastructure management. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fav\u002Fharbor?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fav\u002Fharbor) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fav?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fav) |\n\n\n### Research Papers on Chain-of-Thought Prompting\n\n| Publication Date | Title | 🔗 | Authors | Organization | Technique |\n|------------------|-------|----|---------|--------------|-----------|\n| January 28, 2022 | Chain-of-Thought Prompting Elicits Reasoning in Large Language Models | [🔗](https:\u002F\u002Farxiv.org\u002Fabs\u002F2201.11903) | Jason Wei, et al. | DeepMind | CoT Prompting |\n| March 21, 2022 | Self-Consistency Improves Chain of Thought Reasoning in Language Models | [🔗](https:\u002F\u002Farxiv.org\u002Fabs\u002F2203.11171) | Xuezhi Wang et al. | DeepMind | CoT with Self-Consistency |\n| May 21, 2022 | Least-to-Most Prompting Enables Complex Reasoning in Large Language Models | [🔗](https:\u002F\u002Farxiv.org\u002Fabs\u002F2205.10625) | Denny Zhou et al. | DeepMind | Least-to-Most Prompting |\n| May 21, 2022 | Large Language Models are Zero-Shot Reasoners | [🔗](https:\u002F\u002Farxiv.org\u002Fabs\u002F2205.11916) | Takeshi Kojima, et al. | DeepMind | Zero-shot-CoT |\n| October 6, 2022 | ReAct: Synergizing Reasoning and Acting in Language Models | [🔗](https:\u002F\u002Farxiv.org\u002Fabs\u002F2210.03629) | Shunyu Yao et al. | Princeton University | ReAct |\n| April 1, 2023 | Teaching Large Language Models to Self-Debug | [🔗](https:\u002F\u002Farxiv.org\u002Fabs\u002F2304.05128) | Xiang Lisa Li, et al. | DeepMind, Stanford University | Self-Debugging |\n| May 6, 2023 | Plan-and-Solve Prompting: Improving Zero-Shot Chain-of-Thought Reasoning by Large Language Models | [🔗](https:\u002F\u002Farxiv.org\u002Fabs\u002F2305.04091) | Lei Wang et al. | The Chinese University of Hong Kong, SenseTime Research | Plan-and-Solve Prompting |\n| May 23, 2023 | Let’s Verify Step by Step | [🔗](https:\u002F\u002Farxiv.org\u002Fpdf\u002F2305.20050) | Anya Goyal, et al. | DeepMind | Verification for CoT |\n| October 3, 2023 | Large Language Models Cannot Self-Correct Reasoning Yet | [🔗](https:\u002F\u002Farxiv.org\u002Fabs\u002F2310.01798) | Qingxiu Dong, et al. | The Chinese University of Hong Kong, Huawei Noah's Ark Lab | Self-Correction in LLMs |\n| November 2023 | Universal Self-Consistency for Large Language Model Generation | [🔗](https:\u002F\u002Farxiv.org\u002Fpdf\u002F2311.17311) | Xinyun Chen, Renat Aksitov, Uri Alon, Jie Ren, Kefan Xiao, Pengcheng Yin, Sushant Prakash, Charles Sutton, Xuezhi Wang, Denny Zhou | DeepMind | Universal Self-Consistency |\n| May 17, 2023 | Tree of Thoughts: Deliberate Problem Solving with Large Language Models | [🔗](https:\u002F\u002Fexport.arxiv.org\u002Fabs\u002F2305.10601) | Shunyu Yao, et al. | Princeton University, DeepMind | Tree-of-Thought |\n| February 15, 2024 | Chain-of-Thought Reasoning Without Prompting | [🔗](https:\u002F\u002Farxiv.org\u002Fabs\u002F2402.10200) | Xuezhi Wang, Denny Zhou | DeepMind | Chain-of-Thought Decoding |\n| March 21, 2024 | ChainLM: Empowering Large Language Models with Improved Chain-of-Thought Prompting | [🔗](https:\u002F\u002Farxiv.org\u002Fabs\u002F2403.14238) | Xiaoxue Cheng et al. | Renmin University of China | CoTGenius |\n| June 2024 | Language Agent Tree Search Unifies Reasoning, Acting, and Planning in Language Models | [🔗](https:\u002F\u002Farxiv.org\u002Fpdf\u002F2310.04406) | Andy Zhou, Kai Yan, Michal Shlapentokh-Rothman, Haohan Wang, Yu-Xiong Wang | | Language Agent Tree Search (LATS) |\n| May 2024 | Monte Carlo Tree Search Boosts Reasoning via Iterative Preference Learning | [🔗](https:\u002F\u002Farxiv.org\u002Fpdf\u002F2405.00451) | Yuxi Xie, et al. | National University of Singapore, DeepMind | MCTS |\n| September 18, 2024 | To CoT or not to CoT? Chain-of-thought helps mainly on math and symbolic reasoning | [🔗](https:\u002F\u002Farxiv.org\u002Fabs\u002F2409.12183) | Zayne Sprague, et al. | The University of Texas at Austin, Johns Hopkins University, Princeton University | Meta-analysis of CoT |\n| September 25, 2024 | Chain-of-Thoughtlessness? An Analysis of CoT in Planning | [🔗](https:\u002F\u002Farxiv.org\u002Fabs\u002F2305.12147) | Kaya Stechly, et al. | Arizona State University | Analysis of CoT in Planning |\n| October 18, 2024 | Supervised Chain of Thought | [🔗](https:\u002F\u002Farxiv.org\u002Fabs\u002F2410.14198) | Xiang Zhang, Dujian Ding | University of British Columbia | Supervised Chain of Thought |\n| October 24, 2024 | On examples: A Theoretical Understanding of Chain-of-Thought: Coherent Reasoning and Error-Aware Demonstration | [🔗](https:\u002F\u002Farxiv.org\u002Fabs\u002F2410.16540) | Zhiqiang Hu, et al. | Amazon, Michigan State University | Theoretical Analysis of CoT |\n\n\n\n### CoT Implementations\n\n| Implementation | Link | Author | GitHub Stars | GitHub Followers |\n|----------------|------|---------|--------------|------------------|\n| CoT | [chain-of-thought-hub](https:\u002F\u002Fgithub.com\u002FFranxYao\u002Fchain-of-thought-hub) | Franx Yao | [![Stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FFranxYao\u002Fchain-of-thought-hub?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FFranxYao\u002Fchain-of-thought-hub) | [![Followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FFranxYao?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FFranxYao) |\n| CoT | [optillm](https:\u002F\u002Fgithub.com\u002Fcodelion\u002Foptillm) | Codelion | [![Stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fcodelion\u002Foptillm?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fcodelion\u002Foptillm) | [![Followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fcodelion?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fcodelion) |\n| CoT | [auto-cot](https:\u002F\u002Fgithub.com\u002Famazon-science\u002Fauto-cot) | Amazon Science | [![Stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Famazon-science\u002Fauto-cot?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Famazon-science\u002Fauto-cot) | [![Followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Famazon-science?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Famazon-science) |\n| CoT | [g1](https:\u002F\u002Fgithub.com\u002Fbklieger-groq\u002Fg1) | BKlieger Groq | [![Stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fbklieger-groq\u002Fg1?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fbklieger-groq\u002Fg1) | [![Followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fbklieger-groq?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fbklieger-groq) |\n| Decoding CoT | [optillm\u002Fcot_decoding.py](https:\u002F\u002Fgithub.com\u002Fcodelion\u002Foptillm\u002Fblob\u002Fmain\u002Foptillm\u002Fcot_decoding.py) | Codelion | [![Stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fcodelion\u002Foptillm?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fcodelion\u002Foptillm) | [![Followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fcodelion?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fcodelion) |\n| Tree of Thoughts | [tree-of-thought-llm](https:\u002F\u002Fgithub.com\u002Fprinceton-nlp\u002Ftree-of-thought-llm) | Princeton NLP | [![Stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fprinceton-nlp\u002Ftree-of-thought-llm?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fprinceton-nlp\u002Ftree-of-thought-llm) | [![Followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fprinceton-nlp?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fprinceton-nlp) |\n| Tree of Thoughts | [tree-of-thoughts](https:\u002F\u002Fgithub.com\u002Fkyegomez\u002Ftree-of-thoughts) | Kye Gomez | [![Stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fkyegomez\u002Ftree-of-thoughts?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fkyegomez\u002Ftree-of-thoughts) | [![Followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fkyegomez?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fkyegomez) |\n| Tree of Thoughts | [saplings](https:\u002F\u002Fgithub.com\u002Fshobrook\u002Fsaplings) | Shobrook | [![Stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fshobrook\u002Fsaplings?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fshobrook\u002Fsaplings) | [![Followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fshobrook?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fshobrook) |\n| MCTS | [optillm\u002Fmcts.py](https:\u002F\u002Fgithub.com\u002Fcodelion\u002Foptillm\u002Fblob\u002Fmain\u002Foptillm\u002Fmcts.py) | Codelion | [![Stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fcodelion\u002Foptillm?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fcodelion\u002Foptillm) | [![Followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fcodelion?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fcodelion) |\n| Graph of Thoughts | [graph-of-thoughts](https:\u002F\u002Fgithub.com\u002Fspcl\u002Fgraph-of-thoughts) | SPCL | [![Stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fspcl\u002Fgraph-of-thoughts?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fspcl\u002Fgraph-of-thoughts) | [![Followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fspcl?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fspcl) |\n| Other | [CPO](https:\u002F\u002Fgithub.com\u002Fsail-sg\u002FCPO) | SAIL SG | [![Stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fsail-sg\u002FCPO?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fsail-sg\u002FCPO) | [![Followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fsail-sg?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fsail-sg) |\n| Other | [Everything-of-Thoughts-XoT](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002FEverything-of-Thoughts-XoT) | Microsoft | [![Stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fmicrosoft\u002FEverything-of-Thoughts-XoT?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002FEverything-of-Thoughts-XoT) | [![Followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fmicrosoft?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fmicrosoft) |\n\n### CoT Fine-Tuned Models & Datasets\n\n#### Models\n| Model Name | Author | Size | Link |\n|------------|--------|------|------|\n| CoT-T5-3B | KAIST AI | 3B | [🔗](https:\u002F\u002Fhuggingface.co\u002Fkaist-ai\u002FCoT-T5-3B) |\n| CoT-T5-11B | KAIST AI | 11B | [🔗](https:\u002F\u002Fhuggingface.co\u002Fkaist-ai\u002FCoT-T5-11B) |\n| Llama-3.2V-11B-cot | Xkev | 11B | [🔗](https:\u002F\u002Fhuggingface.co\u002FXkev\u002FLlama-3.2V-11B-cot) |\n| Llama-3.1-8B-Instruct-Reasoner-1o1_v0.3 | Lyte | 8B | [🔗](https:\u002F\u002Fhuggingface.co\u002FLyte\u002FLlama-3.1-8B-Instruct-Reasoner-1o1_v0.3) |\n\n\n#### Datasets\n| Dataset Name | Author | Data Size | Likes | Link |\n|--------------|--------|-----------|-------|------|\n| chain-of-thought-sharegpt | Isaiah Bjork | 7.14k rows | 🌟 8 | [🔗](https:\u002F\u002Fhuggingface.co\u002Fdatasets\u002Fisaiahbjork\u002Fchain-of-thought-sharegpt) |\n| CoT-Collection | KAIST AI | 1.84 million rows | 🌟 122 | [🔗](https:\u002F\u002Fhuggingface.co\u002Fdatasets\u002Fkaist-ai\u002FCoT-Collection?row=4) |\n| Reasoner-1o1-v0.3-HQ | Lyte | 370 rows | 🌟 7 | [🔗](https:\u002F\u002Fhuggingface.co\u002Fdatasets\u002FLyte\u002FReasoner-1o1-v0.3-HQ) |\n| OpenLongCoT-Pretrain | qq8933 | 103k rows | 🌟 86 | [🔗](https:\u002F\u002Fhuggingface.co\u002Fdatasets\u002Fqq8933\u002FOpenLongCoT-Pretrain?row=0) |\n\n\n### Learning Resources\n\n| Tool | Organization | Description | Open Source | GitHub |\n|------|--------------|-------------|-------------|--------|\n| [awesome-cursorrules](https:\u002F\u002Fgithub.com\u002FPatrickJS\u002Fawesome-cursorrules) | PatrickJS | A curated list of resources and guides on cursorrules. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FPatrickJS\u002Fawesome-cursorrules?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FPatrickJS\u002Fawesome-cursorrules) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FPatrickJS?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FPatrickJS) |\n| [ai-engineering-hub](https:\u002F\u002Fgithub.com\u002Fpatchy631\u002Fai-engineering-hub) | patchy631 | A hub of AI engineering learning resources, tutorials, and best practices. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fpatchy631\u002Fai-engineering-hub?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fpatchy631\u002Fai-engineering-hub) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fpatchy631?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fpatchy631) |\n| [GenAI_Agents](https:\u002F\u002Fgithub.com\u002FNirDiamant\u002FGenAI_Agents) | NirDiamant | Resources and examples for building Generative AI Agents. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FNirDiamant\u002FGenAI_Agents?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FNirDiamant\u002FGenAI_Agents) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FNirDiamant?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FNirDiamant) |\n| [learn-agentic-ai](https:\u002F\u002Fgithub.com\u002Fpanaversity\u002Flearn-agentic-ai) | panaversity | Learning materials for understanding and building agentic AI. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fpanaversity\u002Flearn-agentic-ai?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fpanaversity\u002Flearn-agentic-ai) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fpanaversity?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fpanaversity) |\n| [awesome-generative-ai](https:\u002F\u002Fgithub.com\u002Fsteven2358\u002Fawesome-generative-ai) | steven2358 | A curated list of generative AI resources and projects. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fsteven2358\u002Fawesome-generative-ai?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fsteven2358\u002Fawesome-generative-ai) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fsteven2358?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fsteven2358) |\n| [awesome-mcp-servers](https:\u002F\u002Fgithub.com\u002Fpunkpeye\u002Fawesome-mcp-servers) | punkpeye | A curated collection of awesome MCP servers resources. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fpunkpeye\u002Fawesome-mcp-servers?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fpunkpeye\u002Fawesome-mcp-servers) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fpunkpeye?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fpunkpeye) |\n| [GenAI-Showcase](https:\u002F\u002Fgithub.com\u002Fmongodb-developer\u002FGenAI-Showcase) | mongodb-developer | A showcase of innovative Generative AI projects. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fmongodb-developer\u002FGenAI-Showcase?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fmongodb-developer\u002FGenAI-Showcase) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fmongodb-developer?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fmongodb-developer) |\n| [well-architected-iac-analyzer](https:\u002F\u002Fgithub.com\u002Faws-samples\u002Fwell-architected-iac-analyzer) | aws-samples | A tool to analyze and ensure well-architected Infrastructure as Code practices. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Faws-samples\u002Fwell-architected-iac-analyzer?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Faws-samples\u002Fwell-architected-iac-analyzer) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Faws-samples?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Faws-samples) |\n| [llama-cookbook](https:\u002F\u002Fgithub.com\u002Fmeta-llama\u002Fllama-cookbook) | meta-llama | A collection of recipes and guides for working with LLaMA models. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fmeta-llama\u002Fllama-cookbook?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fmeta-llama\u002Fllama-cookbook) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fmeta-llama?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fmeta-llama) |\n| [optillm](https:\u002F\u002Fgithub.com\u002Fcodelion\u002Foptillm) | codelion | Resources for optimizing LLM usage and performance. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fcodelion\u002Foptillm?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fcodelion\u002Foptillm) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fcodelion?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fcodelion) |\n| [cursor.directory](https:\u002F\u002Fgithub.com\u002Fpontusab\u002Fcursor.directory) | pontusab | A directory of tools and resources related to cursor-based workflows. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fpontusab\u002Fcursor.directory?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fpontusab\u002Fcursor.directory) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fpontusab?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fpontusab) |\n| [GenAI_Agents](https:\u002F\u002Fgithub.com\u002FNirDiamant\u002FGenAI_Agents) | NirDiamant | A curated collection of generative AI agents and related tools. | [![GitHub Repo stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FNirDiamant\u002FGenAI_Agents?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FNirDiamant\u002FGenAI_Agents) | [![GitHub followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FNirDiamant?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FNirDiamant) |\n\n\n","# 🌟 令人惊叹的 LLM 资源\n\n精选的 LLM 资源合集。💡✨\n\n**🌐 更新日期：2025年6月22日**\n\n### “无服务器”托管私有\u002F开源模型\n\n| 平台\u002F工具 | 成立时间 | 缩放时间 | 开源 🔓 | GitHub | 入门 | 一键部署 | 开发体验 | 免费 tier |\n| ----------------------------| -------- | -------------| -------------- | ----------|----------------|---------------|-------------------------|--------------|\n| [Baseten](https:\u002F\u002Fwww.baseten.com\u002F) | 2019 | > 15 分钟 | 🔴 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fbasetenlabs?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fbasetenlabs) | [指南](https:\u002F\u002Fdocs.baseten.co\u002Fdeploy\u002Fguides\u002Fprivate-model) | 🟡 | 👍 | $30 |\n| [Modal](https:\u002F\u002Fmodal.com\u002F) | 2021 | \u003C 1 分钟 | 🔴 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fmodal-labs?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fmodal-labs) | [辅助脚本](https:\u002F\u002Fgithub.com\u002Filsilfverskiold\u002FAwesome-LLM-Resources-List\u002Ftree\u002Fmain\u002Fhelpers\u002Fscripts\u002Fmodal) | ❌ | 👍 | $30\u002F月 |\n| [HF Endpoints](https:\u002F\u002Fui.endpoints.huggingface.co\u002F) | 2023 | > 15 分钟 | 🔴 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fhuggingface?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fhuggingface) | 无需额外步骤 | ✅ | 😓 | ❌ |\n| [Replicate](https:\u002F\u002Freplicate.com\u002F) | 2019 | \u003C 1 分钟 | 🔴 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Freplicate?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Freplicate) | [指南](https:\u002F\u002Freplicate.com\u002Fdocs\u002Fguides\u002Fpush-a-transformers-model) | 🟡 | 🤷 | ❌ |\n| [Sagemaker (无服务器)](https:\u002F\u002Faws.amazon.com\u002Fsagemaker\u002F) | 2017 | 不适用 | 🔴 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Faws?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Faws\u002Famazon-sagemaker-examples) | 不适用 | ❌ | ❌ | 300,000 秒 |\n| [Lambda w\u002F EFS (AWS)](https:\u002F\u002Faws.amazon.com\u002Fpm\u002Flambda\u002F) | 2014 | \u003C 1 分钟 | 🔴 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fawsdocs?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fawsdocs\u002Faws-lambda-developer-guide) | [指南](https:\u002F\u002Faws.amazon.com\u002Fblogs\u002Fcompute\u002Fhosting-hugging-face-models-on-aws-lambda\u002F) | ❌ | ❌ | ✅ |\n| [RunPod Serverless](https:\u002F\u002Fwww.runpod.io\u002Fserverless-gpu) | 2022 | > 30 秒 | 🔴 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Frunpod?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Frunpod) | 不适用 | ❌ | 🤷 | ❌ |\n| [BentoML](https:\u002F\u002Fwww.bentoml.com\u002F) | 2019 | > 5 分钟 | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fbentoml\u002FBentoML?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fbentoml\u002FBentoML) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fbentoml?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fbentoml) | [图库](https:\u002F\u002Fwww.bentoml.com\u002Fgallery) | 🟡 | 👍 | 🆓 $10 |\n\n毋庸置疑，这些平台通常还能胜任更多任务，而不仅仅是 LLM 的服务。\n\n### 🧮 无服务器计算定价与限制——Lambda vs Modal（CPU）\n\n| 平台 | 💵 计算单位 | 📥 每次请求费用 | 🆓 免费 tier | ⏱️ 最大超时 | 🚦 并发限制 |\n|--------------------|----------------------------------------------------|----------------------------------------------------|------------------------------------------------------|------------------------------------------|-------------------------------------------------------------------|\n| **AWS Lambda + API GW**| GB-秒 @ $0.000016667 | $0.20\u002FM Lambda + $1.00\u002FM HTTP API 调用 | 100 万 请求 + 40 万 GB-秒\u002F月（12 个月）+ 100 万 API 调用\u002F月 | 15 分钟 | 每个区域 1,000 个（可申请更多） |\n| **Modal** | CPU-秒 @ $0.0000131 + GiB-秒 @ $0.00000222 | ❌ 无每次请求费用 | $30\u002F月 计算积分（Starter 套餐） | 函数：24 小时 ⎮ HTTP：150 秒 → 重定向至 303 页面 | Starter：100 容器 \u002F 200 请求\u002F秒 ⎮ Team：1,000 容器 |\n\n\n### 通过 API 获取现成的开源模型：\n\n| 平台\u002F工具 | 发布时间 | GitHub |\n| --------------------------------------- | -------- | ----------- |\n| [Together.ai](https:\u002F\u002FTogether.ai) | 未注明 | 🔴 |\n| [Fireworks.ai](https:\u002F\u002FFireworks.ai) | 未注明 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Ffw-ai?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Ffw-ai) |\n| [Replicate](https:\u002F\u002Freplicate.com\u002F) | 2019 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Freplicate?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Freplicate) |\n| [Groq](https:\u002F\u002Fgroq.com\u002F) | 未注明 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fgroq?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fgroq) |\n| [DeepInfra](https:\u002F\u002Fdeepinfra.com\u002F) | 未注明 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fdeepinfra?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fdeepinfra) |\n| [Bedrock](https:\u002F\u002Faws.amazon.com\u002Fbedrock) | 未注明 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Faws-samples?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Faws-samples\u002Famazon-bedrock-workshop) |\n| [Lepton](https:\u002F\u002Fwww.lepton.ai\u002F) | 未注明 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fleptonai?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fleptonai) |\n| [Fal.ai](https:\u002F\u002Ffal.ai\u002F) | 未注明 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Ffal-ai?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Ffal-ai) |\n| [VertexAI](https:\u002F\u002Fcloud.google.com\u002Fvertex-ai) | 未注明 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FGoogleCloudPlatform?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FGoogleCloudPlatform\u002Fvertex-ai-samples) |\n\n### 本地推理\n\n| 框架 | 浏览器聊天 🖥️ | 组织 | 开源 | GitHub |\n|-----------------------------------------------|------------------|-------------------|-------------|--------|\n| [Llama.cpp](https:\u002F\u002Fgithub.com\u002Fggerganov\u002Fllama.cpp) | ❌ | ggerganov | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fggerganov\u002Fllama.cpp?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fggerganov\u002Fllama.cpp) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fggerganov?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fggerganov) |\n| [Ollama](https:\u002F\u002Follama.com\u002F) | ❌ | Ollama | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Follama\u002Follama?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Follama\u002Follama) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Follama?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Follama) |\n| [gpt4all](https:\u002F\u002Fwww.nomic.ai\u002Fgpt4all) | ✅ | Nomic.ai | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fnomic-ai\u002Fgpt4all?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fnomic-ai\u002Fgpt4all) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fnomic-ai?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fnomic-ai) |\n| [LMStudio](https:\u002F\u002Flmstudio.ai\u002F) | ✅ | LMStudio AI | 🔴 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Flmstudio-ai?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Flmstudio-ai) |\n| [OpenLLM](https:\u002F\u002Fgithub.com\u002Fbentoml\u002FOpenLLM) | ✅ | BentoML | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fbentoml\u002FOpenLLM?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fbentoml\u002FOpenLLM) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fbentoml?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fbentoml) |\n\n### LLM 服务框架\n\n| 框架 | 开源 | GitHub |\n|----------------------------------------------------------------|---------------------------------------------------------------------------------------------------------------------|------------------------------------------------------------------------------------------|\n| [vLLM](https:\u002F\u002Fgithub.com\u002Fvllm-project\u002Fvllm) | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fvllm-project\u002Fvllm?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fvllm-project\u002Fvllm) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fvllm-project?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fvllm-project) |\n| [OpenLLM](https:\u002F\u002Fgithub.com\u002Fbentoml\u002FOpenLLM) | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fbentoml\u002FOpenLLM?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fbentoml\u002FOpenLLM) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fbentoml?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fbentoml) |\n| [TGI (文本生成推理)](https:\u002F\u002Fgithub.com\u002Fhuggingface\u002Ftext-generation-inference) | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fhuggingface\u002Ftext-generation-inference?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fhuggingface\u002Ftext-generation-inference) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fhuggingface?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fhuggingface) |\n| [TensorRT LLM](https:\u002F\u002Fdocs.nvidia.com\u002Ftensorrt-llm\u002Findex.html) | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FNVIDIA\u002FTensorRT-LLM?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FNVIDIA\u002FTensorRT-LLM) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FNVIDIA?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FNVIDIA) |\n| [Ray Serve](https:\u002F\u002Fdocs.ray.io\u002Fen\u002Flatest\u002Fserve\u002Findex.html) | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fray-project\u002Fray?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fray-project\u002Fray) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fray-project?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fray-project) |\n| [LMDeploy](https:\u002F\u002Fgithub.com\u002FInternLM\u002Flmdeploy) | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FInternLM\u002Flmdeploy?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FInternLM\u002Flmdeploy) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FInternLM?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FInternLM) |\n| [Ollama](https:\u002F\u002Fgithub.com\u002Follama\u002Follama) | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Follama\u002Follama?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Follama\u002Follama) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Follama?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Follama) |\n| [MLC-LLM](https:\u002F\u002Fgithub.com\u002Fmlc-ai\u002Fmlc-llm) | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fmlc-ai\u002Fmlc-llm?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fmlc-ai\u002Fmlc-llm) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fmlc-ai?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fmlc-ai) |\n\n### 构建开源大语言模型Web聊天界面\n\n| 工具 | 组织 | 描述 | 开源 | GitHub |\n|-------------------------------------------------|--------------------|-------------------------------------------------------------------------------------------------|-----------------------------------------------------------------------------------------------------------------------------|------------------------------------------------------------------------------------------|\n| [Text Generation WebUI](https:\u002F\u002Fgithub.com\u002Foobabooga\u002Ftext-generation-webui) | oobabooga | 一个用于大语言模型的Gradio Web界面。 | [![GitHub仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Foobabooga\u002Ftext-generation-webui?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Foobabooga\u002Ftext-generation-webui) | [![GitHub关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Foobabooga?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Foobabooga) |\n| [Jan AI](https:\u002F\u002Fjan.ai\u002F) | Jan HQ | 一个完全在您本地电脑上离线运行的ChatGPT开源替代方案。支持多种引擎（llama.cpp、TensorRT-LLM）。 | [![GitHub仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fjanhq\u002Fjan?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fjanhq\u002Fjan) | [![GitHub关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fjanhq?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fjanhq) |\n| [AnythingLLM](https:\u002F\u002Fgithub.com\u002FMintplex-Labs\u002Fanything-llm) | Mintplex Labs | 一款集成了RAG、AI代理等功能的一体化桌面及Docker AI应用。 | [![GitHub仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FMintplex-Labs\u002Fanything-llm?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FMintplex-Labs\u002Fanything-llm) | [![GitHub关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FMintplex-Labs?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FMintplex-Labs) |\n| [Superagent](https:\u002F\u002Fgithub.com\u002Fsuperagent-ai\u002Fsuperagent) | Superagent AI | 允许开发者使用LLM和RAG将强大的AI助手集成到他们的应用中。 | [![GitHub仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fsuperagent-ai\u002Fsuperagent?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fsuperagent-ai\u002Fsuperagent) | [![GitHub关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fsuperagent-ai?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fsuperagent-ai) |\n| [Bionic-GPT](https:\u002F\u002Fbionic-gpt.com\u002F) | Bionic GPT | 一款ChatGPT替代品，提供生成式AI优势的同时，严格保障数据隐私。 | [![GitHub仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fbionic-gpt\u002Fbionic-gpt?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fbionic-gpt\u002Fbionic-gpt) | [![GitHub关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fbionic-gpt?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fbionic-gpt) |\n| [Open WebUI](https:\u002F\u002Fgithub.com\u002Fopen-webui\u002Fopen-webui) | Open WebUI | 一个用户友好的Web界面，用于与大语言模型（LLMs）交互。 | [![GitHub仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fopen-webui\u002Fopen-webui?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fopen-webui\u002Fopen-webui) | [![GitHub关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fopen-webui?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fopen-webui) |\n| [Xyne](https:\u002F\u002Fgithub.com\u002Fxynehq\u002Fxyne) | xynehq | 一个简洁、极简的Web聊天界面，用于与大语言模型交互。 | [![GitHub仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fxynehq\u002Fxyne?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fxynehq\u002Fxyne) | [![GitHub关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fxynehq?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fxynehq) |\n| [Assistant UI](https:\u002F\u002Fgithub.com\u002Fassistant-ui\u002Fassistant-ui) | assistant-ui | 一个开源的类似ChatGPT的界面，设计简洁且响应迅速。 | [![GitHub仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fassistant-ui\u002Fassistant-ui?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fassistant-ui\u002Fassistant-ui) | [![GitHub关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fassistant-ui?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fassistant-ui) |\n| [Scira](https:\u002F\u002Fgithub.com\u002Fzaidmukaddam\u002Fscira) | zaidmukaddam | 一个基于AI的搜索界面，利用LLM提供智能搜索结果。 | [![GitHub仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fzaidmukaddam\u002Fscira?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fzaidmukaddam\u002Fscira) | [![GitHub关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fzaidmukaddam?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fzaidmukaddam) |\n| [Onyx](https:\u002F\u002Fgithub.com\u002Fonyx-dot-app\u002Fonyx) | onyx-dot-app | 一个可定制且可扩展的Web聊天界面，用于与大语言模型交互。 | [![GitHub仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fonyx-dot-app\u002Fonyx?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fonyx-dot-app\u002Fonyx) | [![GitHub关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fonyx-dot-app?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fonyx-dot-app) |\n| [NextChat](https:\u002F\u002Fgithub.com\u002FChatGPTNextWeb\u002FNextChat) | ChatGPTNextWeb | 一个基于Next.js的开源ChatGPT克隆，用于无缝的Web交互。 | [![GitHub仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FChatGPTNextWeb\u002FNextChat?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FChatGPTNextWeb\u002FNextChat) | [![GitHub关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FChatGPTNextWeb?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FChatGPTNextWeb) |\n\n### 租用 GPU（微调、部署、训练）\n\n| 平台 | 模板 | 是否适合初学者 | GitHub |\n|-----------------------------------------------|--------------------------|-------------------|--------|\n| [Brev.dev](https:\u002F\u002Fwww.brev.dev\u002F) | 微调 | ❌ | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fbrevdev?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fbrevdev) |\n| [Modal](https:\u002F\u002Fmodal.com\u002F) | 微调 | ❌ | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fmodal-labs?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fmodal-labs) |\n| [Hyperbolic AI](https:\u002F\u002Fhyperbolic.xyz\u002F) | 无 | ❌ | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FHyperbolicLabs?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FHyperbolicLabs) |\n| [RunPod](https:\u002F\u002Fwww.runpod.io\u002F) | 无 | ❌ | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Frunpod?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Frunpod) |\n| [Paperspace](https:\u002F\u002Fwww.paperspace.com\u002F) | 微调 | ✅ | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FPaperspace?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FPaperspace) |\n| [Colab](https:\u002F\u002Fcolab.research.google.com\u002F) | 仅小型模型 | ✅ | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fgooglecolab?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fgooglecolab) |\n\n\n### 无需代码的 UI 微调工具\n\n| 工具 | 是否适合初学者 | 开源 | GitHub |\n|------------------------------------------------|-------------------|-------------|--------|\n| [Together.ai](https:\u002F\u002Fwww.together.ai\u002Fproducts#fine-tuning) | ✅ | ❌ | N\u002FA |\n| [Hugging Face AutoTrain](https:\u002F\u002Fhuggingface.co\u002Fautotrain) | ✅ | ❌ | [![GitHub 仓库星级](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fhuggingface\u002Fautotrain-advanced?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fhuggingface\u002Fautotrain-advanced) |\n| [AutoML](https:\u002F\u002Fgithub.com\u002Fmljar\u002Fautoml-app) | ❌ | ✅ | [![GitHub 仓库星级](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fmljar\u002Fautoml-app?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fmljar\u002Fautoml-app) |\n| [LLaMA-Factory](https:\u002F\u002Fgithub.com\u002Fhiyouga\u002FLLaMA-Factory) | ❌ | ✅ | [![GitHub 仓库星级](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fhiyouga\u002FLLaMA-Factory?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fhiyouga\u002FLLaMA-Factory) |\n| [H2O LLM Studio](https:\u002F\u002Fgithub.com\u002Fh2oai\u002Fh2o-llmstudio) | ✅ | ✅ | [![GitHub 仓库星级](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fh2oai\u002Fh2o-llmstudio?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fh2oai\u002Fh2o-llmstudio) |\n\n### 微调框架\n\n| 框架 | 开源 | GitHub |\n|-------------------------------------------|-------------|--------|\n| [Axolotl](https:\u002F\u002Faxolotl.ai\u002F) | [![GitHub 仓库星级](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Faxolotl-ai-cloud\u002Faxolotl?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Faxolotl-ai-cloud\u002Faxolotl) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Faxolotl-ai-cloud?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Faxolotl-ai-cloud) |\n| [Unsloth](https:\u002F\u002Funsloth.ai\u002F) | [![GitHub 仓库星级](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Funslothai\u002Funsloth?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Funslothai\u002Funsloth) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Funslothai?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Funslothai) |\n\n\n### 操作系统代理\u002FAI 工作流\n\n| 框架 | 开源 | 初学者友好 | 发布年份 | GitHub |\n|--------------------------------------------------|-----------------------------------------------------------------------------------------------------------------------------|-------------------|----------|------------------------------------------------------------------------------------------|\n| [LangChain](https:\u002F\u002Fwww.langchain.com\u002F) | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Flangchain-ai\u002Flangchain?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Flangchain-ai\u002Flangchain) | ✅ | 2022 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Flangchain-ai?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Flangchain-ai) |\n| [LlamaIndex](https:\u002F\u002Fdocs.llamaindex.ai\u002Fen\u002Fstable\u002Fuse_cases\u002Fagents\u002F) | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Frun-llama\u002Fllama_index?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Frun-llama\u002Fllama_index) | ❌ | 2023 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Frun-llama?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Frun-llama) |\n| [Swarms](https:\u002F\u002Fswarms.world\u002F) | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fkyegomez\u002Fswarms?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fkyegomez\u002Fswarms) | ❌ | 2023 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fkyegomez?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fkyegomez) |\n| [CrewAI](https:\u002F\u002Fwww.crewai.com\u002F) | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fcrewaiinc\u002Fcrewai?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fcrewaiinc\u002Fcrewai) | ✅ | 2023 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fcrewaiinc?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fcrewaiinc) |\n| [Autogen](https:\u002F\u002Fmicrosoft.github.io\u002Fautogen\u002F0.2\u002F) | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fmicrosoft\u002Fautogen?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fautogen) | ✅ | 2023 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fmicrosoft?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fmicrosoft) |\n| [AutoChain](https:\u002F\u002Fautochain.forethought.ai\u002F) | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FForethought-Technologies\u002FAutoChain?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FForethought-Technologies\u002FAutoChain) | ❌ | 2023 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FForethought-Technologies?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FForethought-Technologies) |\n| [SuperAGI](https:\u002F\u002Fsuperagi.com\u002F) | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FTransformerOptimus\u002FSuperAGI?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FTransformerOptimus\u002FSuperAGI) | ❌ | 2023 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FTransformerOptimus?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FTransformerOptimus) |\n| [AILegion](https:\u002F\u002Fgithub.com\u002Feumemic\u002Fai-legion) | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Feumemic\u002Fai-legion?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Feumemic\u002Fai-legion) | ❌ | 2023 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Feumemic?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Feumemic) |\n| [MemGPT (Letta)](https:\u002F\u002Fwww.letta.com\u002F) | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fcpacker\u002FMemGPT?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fcpacker\u002FMemGPT) | ❌ | 2023 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fcpacker?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fcpacker) |\n| [uAgents](https:\u002F\u002Fpypi.org\u002Fproject\u002Fuagents\u002F) | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Ffetchai\u002FuAgents?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Ffetchai\u002FuAgents) | ❌ | 2023 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Ffetchai?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Ffetchai) |\n| [AGiXT](https:\u002F\u002Fgithub.com\u002FJosh-XT\u002FAGiXT) | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FJosh-XT\u002FAGiXT?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FJosh-XT\u002FAGiXT) | ❌ | 2023 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FJosh-XT?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FJosh-XT) |\n| [Dify](https:\u002F\u002Fdify.ai\u002F) | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Flanggenius\u002Fdify?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Flanggenius\u002Fdify) | ✅ | 2024 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Flanggenius?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Flanggenius) |\n| [TaskingAI](https:\u002F\u002Fwww.tasking.ai\u002F) | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FTaskingAI\u002FTaskingAI?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FTaskingAI\u002FTaskingAI) | ✅ | 2024 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FTaskingAI?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FTaskingAI) |\n| [Bee Agent Framework](https:\u002F\u002Fi-am-bee.github.io\u002Fbee-agent-framework\u002F#\u002F) | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fi-am-bee\u002Fbee-agent-framework?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fi-am-bee\u002Fbee-agent-framework) | ❌ | 2024 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fi-am-bee?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fi-am-bee) |\n| [Swarms](https:\u002F\u002Fswarms.world\u002F) | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fkyegomez\u002Fswarms?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fkyegomez\u002Fswarms) | ❌ | 2024 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fkyegomez?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fkyegomez) |\n| [IoA](https:\u002F\u002Fopenbmb.github.io\u002FIoA\u002F) | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FOpenBMB\u002FIoA?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FOpenBMB\u002FIoA) | ❌ | 2024 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FOpenBMB?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FOpenBMB) |\n| [Upsonic](https:\u002F\u002Fgithub.com\u002FUpsonic\u002FUpsonic) | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FUpsonic\u002FUpsonic?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FUpsonic\u002FUpsonic) | ❌ | 2024 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FUpsonic?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FUpsonic) |\n| [Parlant](https:\u002F\u002Fgithub.com\u002Femcie-co\u002Fparlant) | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Femcie-co\u002Fparlant?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Femcie-co\u002Fparlant) | ❌ | 2024 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Femcie-co?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Femcie-co) |\n| [Rig](https:\u002F\u002Fgithub.com\u002F0xPlaygrounds\u002Frig) | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002F0xPlaygrounds\u002Frig?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002F0xPlaygrounds\u002Frig) | ❌ | 2024 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002F0xPlaygrounds?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002F0xPlaygrounds) |\n| [eliza](https:\u002F\u002Fgithub.com\u002FelizaOS\u002Feliza) | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FelizaOS\u002Feliza?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FelizaOS\u002Feliza) | ✅ | 2024 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FelizaOS?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FelizaOS) |\n| [TensorZero](https:\u002F\u002Fwww.tensorzero.com\u002F) | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Ftensorzero\u002Ftensorzero?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Ftensorzero\u002Ftensorzero) | ❌ | 2024 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Ftensorzero?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Ftensorzero\u002Ftensorzero) |\n| [AgentDock](https:\u002F\u002Fagentdock.ai\u002F) | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FAgentDock\u002FAgentDock?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FAgentDock\u002FAgentDock) | ✅ | 2025 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FAgentDock?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FAgentDock) |\n\n### 顶级代理框架\n\n| 框架 | 开源 | 初学者友好 | 发布年份 | GitHub |\n|-----------------------------------------------------------|-----------------------------------------------------------------------------------------------------------------------------|-------------------|----------|------------------------------------------------------------------------------------------|\n| [LangGraph](https:\u002F\u002Fwww.langchain.com\u002Flanggraph) | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Flangchain-ai\u002Flanggraph?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Flangchain-ai\u002Flanggraph) | ❌ | 2023 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Flangchain-ai?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Flangchain-ai) |\n| [Flowise](https:\u002F\u002Fflowiseai.com\u002F) | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FFlowiseAI\u002FFlowise?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FFlowiseAI\u002FFlowise) | ✅ | 2023 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FFlowiseAI?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FFlowiseAI) |\n| [Langroid](https:\u002F\u002Flangroid.github.io\u002Flangroid\u002F) | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Flangroid\u002Flangroid?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Flangroid\u002Flangroid) | ❌ | 2023 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Flangroid?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Flangroid) |\n| [smolagents](https:\u002F\u002Fgithub.com\u002Fhuggingface\u002Fsmolagents) | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fhuggingface\u002Fsmolagents?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fhuggingface\u002Fsmolagents) | ❌ | 2024 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fhuggingface?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fhuggingface) |\n| [Semantic Kernel](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fsemantic-kernel) | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fmicrosoft\u002Fsemantic-kernel?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fsemantic-kernel) | ❌ | 2023 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fmicrosoft?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fmicrosoft) |\n| [Atomic Agents](https:\u002F\u002Fgithub.com\u002FBrainBlend-AI\u002Fatomic-agents) | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FBrainBlend-AI\u002Fatomic-agents?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FBrainBlend-AI\u002Fatomic-agents) | ❌ | 2024 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FBrainBlend-AI?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FBrainBlend-AI) |\n| [Agno](https:\u002F\u002Fgithub.com\u002Fagno-agi\u002Fagno) | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fpydantic\u002Fpydantic-ai?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fagno-agi\u002Fagno) | ✅ | 2024 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fpydantic?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fagno-agi) |\n| [PydanticAI](https:\u002F\u002Fgithub.com\u002Fpydantic\u002Fpydantic-ai) | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fpydantic\u002Fpydantic-ai?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fpydantic\u002Fpydantic-ai) | ❌ | 2024 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fpydantic?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fpydantic) |\n| [Mastra](https:\u002F\u002Fgithub.com\u002Fmastra-ai\u002Fmastra) | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fmastra-ai\u002Fmastra?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fmastra-ai\u002Fmastra) | ✅ | 2025 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fmastra-ai?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fmastra-ai) |\n\n### 代理框架：核心能力\n| 框架 | 内存与 RAG | 多模态性 | 多智能体支持 | 可观测性 |\n|----------------|-----------------------------------------------|-----------------------------------------------|-----------------------------------------------|-------------------------------------------------|\n| **AgentDock** | 内置 RAG 系统；知识库集成 | 🟢 多模态（文本、语音、工具、API） | 视觉化工作流编排与智能体链 | 全面的 LLM 追踪与信用记录跟踪 |\n| **Agno** | 集成内存与向量数据库\u002FRAG | 🟢 原生（文本、图像、音频、视频） | 监督者-工作者角色 | 内置云仪表盘\u002F日志系统 |\n| **LangGraph** | 持久化状态；易于外部集成 | 🔸 主要为文本；可通过节点扩展 | 层次化编排 | LangSmith 集成与图编辑器 |\n| **SmolAgents** | 内置短期记忆；可自定义长期记忆 | 🔸 通过 VLM 实现视觉智能体 | 模块化多智能体组合 | 较少；依赖外部日志系统 |\n| **Mastra** | 持久化工作流；原生 RAG 流程 | 🟢 通过集成实现多模态 | 原生多智能体工作流 | 内置 OpenTelemetry 仪表盘 |\n| **Pydantic AI**| 基于依赖注入的内存与 RAG 集成 | 🔸 以文本为主；可通过自定义 DI 实现多模态 | 类型安全的手动编排 | 有限；使用 Python 日志\u002FOpenTelemetry |\n| **Atomic Agents**| 每个智能体独立的内存与 RAG（向量数据库） | 🟢 原生多模态 | 显式的工作流串联 | 较少；建议使用外部监控工具 |\n| **Autogen** | 内置短期记忆；可接入外部长期记忆 | 🔸 以文本为主；可扩展 | 自发的、自由形式的合作 | 中等；内部日志，无仪表盘 |\n| **CrewAI** | 带状态的记忆与团队式 RAG | 🟢 多种模态（文本、图像等） | 监督者主导的多团队工作流 | 集成的仪表盘用于日志与监控 |\n\n请参阅此 [Google 表格](https:\u002F\u002Fdocs.google.com\u002Fspreadsheets\u002Fd\u002F1zjcww1w0vARZz9Z6GDxNMp-PKyg7iRyNYAnDo59HjzI\u002Fedit?usp=sharing)，其中包含更多列。\n\n### 视觉 AI 代理构建工具\n\n| 工具 | 组织 | 描述 | 开源 | GitHub |\n|------|--------------|-------------|-------------|--------|\n| [Rivet](https:\u002F\u002Fgithub.com\u002FIronclad\u002Frivet) | Ironclad | 用于设计和部署 AI 代理工作流的可视化构建工具。 | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FIronclad\u002Frivet?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FIronclad\u002Frivet) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FIronclad?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FIronclad) |\n| [PySpur](https:\u002F\u002Fgithub.com\u002FPySpur-Dev\u002Fpyspur) | PySpur-Dev | 一款无缝构建和可视化 AI 代理的工具。 | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FPySpur-Dev\u002Fpyspur?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FPySpur-Dev\u002Fpyspur) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FPySpur-Dev?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FPySpur-Dev) |\n| [Flowise](https:\u002F\u002Fgithub.com\u002FFlowiseAI\u002FFlowise) | FlowiseAI | 一个无代码、可视化的平台，用于设计 AI 代理工作流。 | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FFlowiseAI\u002FFlowise?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FFlowiseAI\u002FFlowise) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FFlowiseAI?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FFlowiseAI) |\n\n### 💬 代理系统模型调用定价（纯文本，2000 tokens 输入，100 tokens 输出，统一费率）\n\n| 模型 | 💵 每次调用价格 | 💯 100 次调用 | 🧮 1,000 次调用 | 🔁 30,000 次调用 |\n|---------------------|------------:|-------------:|---------------:|----------------:|\n| **Gemini Flash 2.0**| $0.00000 | $0.02 | $0.24 | $7.20 |\n| **GPT-4o mini** | $0.00144 | $0.14 | $1.44 | $43.20 |\n| **GPT-4.1** | $0.00480 | $0.48 | $4.80 | $144.00 |\n| **Gemini Pro 2.5** | $0.00350 | $0.35 | $3.50 | $105.00 |\n| **Claude Haiku 3.5**| $0.00200 | $0.20 | $2.00 | $60.00 |\n| **Claude Sonnet 4** | $0.00750 | $0.75 | $7.50 | $225.00 |\n| **GPT-4o** | $0.01200 | $1.20 | $12.00 | $360.00 |\n| **OpenAI o3** | $0.02400 | $2.40 | $24.00 | $720.00 |\n\n### 代理相关工具（用于“构建”）\n\n| 工具 | 组织 | 描述 | 开源 | GitHub |\n|------|--------------|-------------|-------------|--------|\n| [browser-use](https:\u002F\u002Fgithub.com\u002Fbrowser-use\u002Fbrowser-use) | browser-use | 将浏览器功能集成到代理工作流中。 | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fbrowser-use\u002Fbrowser-use?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fbrowser-use\u002Fbrowser-use) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fbrowser-use?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fbrowser-use) |\n| [code2prompt](https:\u002F\u002Fgithub.com\u002Fmufeedvh\u002Fcode2prompt) | mufeedvh | 将代码片段转换为可用于开发的可操作提示。 | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fmufeedvh\u002Fcode2prompt?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fmufeedvh\u002Fcode2prompt) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fmufeedvh?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fmufeedvh) |\n| [note-gen](https:\u002F\u002Fgithub.com\u002Fcodexu\u002Fnote-gen) | codexu | 自动从你的代码中生成笔记和文档。 | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fcodexu\u002Fnote-gen?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fcodexu\u002Fnote-gen) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fcodexu?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fcodexu) |\n| [refly](https:\u002F\u002Fgithub.com\u002Frefly-ai\u002Frefly) | refly-ai | 自动化代码重构和提示生成任务。 | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Frefly-ai\u002Frefly?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Frefly-ai\u002Frefly) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Frefly-ai?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Frefly-ai) |\n| [potpie](https:\u002F\u002Fgithub.com\u002Fpotpie-ai\u002Fpotpie) | potpie-ai | 一个用于原型设计和构建 AI 代理流水线的工具包。 | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fpotpie-ai\u002Fpotpie?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fpotpie-ai\u002Fpotpie) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fpotpie-ai?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fpotpie-ai) |\n| [AgentStack](https:\u002F\u002Fgithub.com\u002FAgentOps-AI\u002FAgentStack) | AgentOps-AI | 一个用于构建和部署 AI 代理的综合栈。 | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FAgentOps-AI\u002FAgentStack?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FAgentOps-AI\u002FAgentStack) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FAgentOps-AI?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FAgentOps-AI) |\n| [browser](https:\u002F\u002Fgithub.com\u002Flightpanda-io\u002Fbrowser) | lightpanda-io | 一款基于浏览器的工具，专为集成代理功能而设计。 | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Flightpanda-io\u002Fbrowser?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Flightpanda-io\u002Fbrowser) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Flightpanda-io?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Flightpanda-io) |\n| [Memary](https:\u002F\u002Fgithub.com\u002Fkingjulio8238\u002FMemary) | kingjulio8238 | 一个用于在代理工作流中保持上下文的记忆模块。 | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fkingjulio8238\u002FMemary?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fkingjulio8238\u002FMemary) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fkingjulio8238?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fkingjulio8238) |\n| [open-canvas](https:\u002F\u002Fgithub.com\u002Flangchain-ai\u002Fopen-canvas) | langchain-ai | 一个用于使用 LangChain 设计代理工作流的可视化界面。 | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Flangchain-ai\u002Fopen-canvas?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Flangchain-ai\u002Fopen-canvas) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Flangchain-ai?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Flangchain-ai) |\n| [agent-service-toolkit](https:\u002F\u002Fgithub.com\u002FJoshuaC215\u002Fagent-service-toolkit) | JoshuaC215 | 一个用于构建和部署基于代理的服务的工具包。 | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FJoshuaC215\u002Fagent-service-toolkit?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FJoshuaC215\u002Fagent-service-toolkit) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FJoshuaC215?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FJoshuaC215) |\n\n### 虚拟大脑\n\n| 工具 | 组织 | 描述 | 开源 | GitHub |\n|------|--------------|-------------|-------------|--------|\n| [Leon](https:\u002F\u002Fgithub.com\u002Fleon-ai\u002Fleon) | leon-ai | 一个由人工智能驱动的开源个人助理和自动化平台。 | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fleon-ai\u002Fleon?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fleon-ai\u002Fleon) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fleon-ai?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fleon-ai) |\n| [Khoj](https:\u002F\u002Fgithub.com\u002Fkhoj-ai\u002Fkhoj) | khoj-ai | 一个利用人工智能组织和检索知识的虚拟大脑。 | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fkhoj-ai\u002Fkhoj?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fkhoj-ai\u002Fkhoj) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fkhoj-ai?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fkhoj-ai) |\n\n\n### AI 代理\n\n| 框架 | 组织 | 开源 | 发布时间 | GitHub |\n|-----------------------------------------------------|------------------------------|-----------------------------------------------------------------------------------------------------------------------------|----------|------------------------------------------------------------------------------------------|\n| [GPT Engineer](https:\u002F\u002Fgptengineer.app\u002F) | GPT Engineer Org | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fgpt-engineer-org\u002Fgpt-engineer?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fgpt-engineer-org\u002Fgpt-engineer) | 2023 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fgpt-engineer-org?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fgpt-engineer-org) |\n| [XAgent](https:\u002F\u002Fgithub.com\u002FOpenBMB\u002FXAgent) | OpenBMB | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FOpenBMB\u002FXAgent?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FOpenBMB\u002FXAgent) | 2023 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FOpenBMB?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FOpenBMB) |\n| [Bolt.new](https:\u002F\u002Fgithub.com\u002Fstackblitz\u002Fbolt.new) | StackBlitz | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fstackblitz\u002Fbolt.new?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fstackblitz\u002Fbolt.new) | 2023 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fstackblitz?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fstackblitz) |\n| [Goose](https:\u002F\u002Fgithub.com\u002Fblock\u002Fgoose) | Block | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fblock\u002Fgoose?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fblock\u002Fgoose) | 2023 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fblock?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fblock) |\n| [AI Hedge Fund](https:\u002F\u002Fgithub.com\u002Fvirattt\u002Fai-hedge-fund) | virattt | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fvirattt\u002Fai-hedge-fund?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fvirattt\u002Fai-hedge-fund) | 2023 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fvirattt?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fvirattt) |\n| [FinRobot](https:\u002F\u002Fai4finance.org\u002F) | AI4Finance Foundation | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FAI4Finance-Foundation\u002FFinRobot?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FAI4Finance-Foundation\u002FFinRobot) | 2024 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FAI4Finance-Foundation?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FAI4Finance-Foundation) |\n| [STORM](https:\u002F\u002Fstorm.genie.stanford.edu\u002F) | Stanford OVAL | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fstanford-oval\u002Fstorm?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fstanford-oval\u002Fstorm) | 2024 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fstanford-oval?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fstanford-oval) |\n| [Multion](https:\u002F\u002Fwww.multion.ai\u002F) | MULTI-ON | 🔴 | 无 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FMULTI-ON?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FMULTI-ON) |\n| [Minion](https:\u002F\u002Fminion.ai\u002F) | Minion AI | 🔴 | 无 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fminionai?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fminionai) |\n\n### 长期记忆\n\n| 提供商 | 社区 | 成立时间 | GitHub | ⭐ 星数 | 开源 |\n|--------------|------------------------|-------------|--------------------------------------------------------------------------------------------------|---------|-----------------------------| \n| Letta | 💬 活跃开发者社区| 2023年10月 | [![GitHub 粉丝](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fletta-ai?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fletta-ai\u002Fletta) | 1.7万 | ✅ Apache-2.0 |\n| Zep | 🤝 中等规模社区 | 2024年8月 | [![GitHub 粉丝](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fgetzep?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fgetzep\u002Fgraphiti) | 1.16万 | ⚠️ Graphiti CE (Apache-2.0) |\n| MemoRAG | 🧪 小型研究团队| 2024年9月 | [![GitHub 粉丝](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fqhjqhj00?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fqhjqhj00\u002FMemoRAG) | 1,800 | ✅ Apache-2.0 |\n| Memary | 🧠 小众社区 | 2024年4月 | [![GitHub 粉丝](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fkingjulio8238?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fkingjulio8238\u002FMemary) | 2,300 | ✅ MIT |\n| Cognee | 🔄 中等规模 | 2023年8月 | [![GitHub 粉丝](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Ftopoteretes?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Ftopoteretes\u002Fcognee) | 5,800 | ✅ Apache-2.0 |\n| Mem0 | 🚀 快速增长 | 2023年6月 | [![GitHub 粉丝](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fmem0ai?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fmem0ai\u002Fmem0) | 3.52万 | ✅ Apache-2.0 |\n\n#### 记忆功能对比\n\n| 提供商 | 基础架构 | 可选知识图谱 | 自我编辑\u002F自主性 | 循环摘要 | 分类标签 |\n|-----------|---------|-------------|-------------------------|-------------------------------|------------|\n| Letta | 🧮 向量 | ⚠️ 部分支持 | ✅ 支持 | ⚠️ 部分支持（记忆块） | ✅ 支持 |\n| Zep | 🧠 知识图谱 | - | ✅ 支持 | ✅ 自动聊天摘要 | ✅ 支持 |\n| MemoRAG | 🧮 向量 | ❌ 不支持 | ✅ 支持 | ❌ 使用长距离模型 | ❌ 不支持 |\n| Memary | 🧠 知识图谱 | - | ✅ 支持 | ⚠️ 计划推出“回溯”功能 | ✅ 支持 |\n| Cognee | 🧠 知识图谱 | - | ✅ 支持 | ❌ 无自动摘要 | ✅ 支持 |\n| Mem0 | 🧮 向量 | ❌ 不支持 | ✅ 支持 | ❌ 未明确说明 | ✅ 支持 |\n\n#### 企业级安全性（云端使用）\n\n| 提供商 | 企业级安全特性 |\n|-----------|--------------------------------------------------------------------------------------------------------|\n| Mem0 | 🔐 托管服务，支持加密、组织和项目角色管理，符合 GDPR 的数据删除机制。采用 Graphlit 技术（未声明 SOC 2）。 |\n| Letta | ☁️ 支持自托管或托管服务器。用户身份验证及按 ID 分区的内存。基于 Graphlit 架构。公开 SSO 详情缺失。 |\n| Zep | ✅ SOC 2 Type 2 认证。静态与传输中加密、访问控制、JWT 以及删除 API（“被遗忘权”）。 |\n| MemoRAG | 🏠 支持自托管 |\n| Memary | 🏠 支持自托管 |\n| Cognee | 🏠 支持自托管 |\n\n\n#### 按每月消息量定价\n\n| 提供商 | 1千条消息\u002F月 | 1万条消息\u002F月 | 10万条消息\u002F月 | 100万条消息\u002F月 |\n|------------|------------------|---------------------|---------------------------|---------------------------|\n| Mem0 | 🆓 免费 | 🆓 免费–$29 | 💵 $249 | 🏢 企业级（定制） |\n| Letta | 🆓 免费 | 💵 $20 | 💵 $750 | 🏢 企业级（定制） |\n| Zep | 🆓 免费 | 🆓 免费 | 💵 ~ $112.50 | 💵 ~ $1,237 |\n| MemoRAG | 💻 GPU 服务器 (~$150–300\u002F月) | 💻 GPU 服务器 (~$150–300\u002F月) | 💻 多 GPU（$500+） | 🖥️ 集群（$1K+\u002F月） |\n| 自托管 | 🖥️ 小型虚拟机 (~$15\u002F月) | 🖥️ 小型虚拟机 (~$15–20\u002F月) | 🖥️ 中型虚拟机 ($50–$100\u002F月) | 🖥️ 大型虚拟机 ($200+\u002F月) |\n\n### 评估框架及附加组件\n\n| 框架 | 开源 | 初学者友好 | 发布年份 | GitHub |\n| ------------------------------------------------------------------ | ----------------------------------------------------------------------------------------------------------------------------------------------------------------- | ----------------- | -------- | ------------------------------------------------------------------------------------------------------------------------------------------------------ |\n| [TruLens](https:\u002F\u002Fgithub.com\u002Ftruera\u002Ftrulens) | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Ftruera\u002Ftrulens?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Ftruera\u002Ftrulens) | ✅ | 2023 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Ftruera?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Ftruera) |\n| [Promptfoo](https:\u002F\u002Fgithub.com\u002Fpromptfoo\u002Fpromptfoo) | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fpromptfoo\u002Fpromptfoo?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fpromptfoo\u002Fpromptfoo) | ✅ | 2023 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fpromptfoo?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fpromptfoo) |\n| [DeepEval](https:\u002F\u002Fgithub.com\u002Fconfident-ai\u002Fdeepeval) | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fconfident-ai\u002Fdeepeval?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fconfident-ai\u002Fdeepeval) | ✅ | 2024 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fconfident-ai?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fconfident-ai) |\n| [RAGAS](https:\u002F\u002Fgithub.com\u002Fexplodinggradients\u002Fragas) | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fexplodinggradients\u002Fragas?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fexplodinggradients\u002Fragas) | ❌ | 2023 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fexplodinggradients?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fexplodinggradients) |\n| [OpenAI Evals](https:\u002F\u002Fgithub.com\u002Fopenai\u002Fevals) | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fopenai\u002Fevals?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fopenai\u002Fevals) | ❌ | 2023 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fopenai?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fopenai) |\n| [LangChain OpenEvals](https:\u002F\u002Fgithub.com\u002Flangchain-ai\u002Fopenevals) | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Flangchain-ai\u002Fopenevals?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Flangchain-ai\u002Fopenevals) | ✅ | 2025 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Flangchain-ai?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Flangchain-ai) |\n| [LangChain AgentEvals](https:\u002F\u002Fgithub.com\u002Flangchain-ai\u002Fagentevals) | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Flangchain-ai\u002Fagentevals?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Flangchain-ai\u002Fagentevals) | ❌ | 2025 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Flangchain-ai?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Flangchain-ai) |\n| [LlamaIndex Eval](https:\u002F\u002Fgithub.com\u002Frun-llama\u002Fllama_index) | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Frun-llama\u002Fllama_index?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Frun-llama\u002Fllama_index) | ✅ | 2023 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Frun-llama?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Frun-llama) |\n\n### 评估框架：核心差异\n\n| 框架 | Pytest \u002F CLI 运行器 | 预置指标 | 合成数据生成 | 离线评判 | 模型无关 | 安全性 | 红队测试 | **自定义指标（设置速度）** |\n| ------------------- | --------------------- | ------------------ | ---------------------------- | ------------- | -------------- | ------ | -------- | -------------------------------------------------- |\n| **DeepEval** | 🟢 `deepeval test` | **40 +** | 🟢 `deepeval create-dataset` | 🟢 | 🟢 | 🟢 | 🟢 | 🟢 **G-Eval 构建器 — 几分钟（单个函数）** |\n| **RAGAS** | ✖（脚本断言） | 6 核心 RAG + 🔸 | 🟢 基于知识图谱的问答生成 | 🟢 | 🟢 | 🔸 自行搭建 | ✖ | 🟢 **`AspectCritic` 单行代码 — 几分钟** |\n| **MLflow Evaluate** | ✖（`mlflow.evaluate`） | 3-4 | ✖ 用户自备 | 🔸 可能 | 🔸 | 🟢 | ✖ | 🟢 **子类评分器 — 几行代码，约一小时** |\n| **OpenAI Evals** | 🟢 CLI 协调工具 | ~10 个模板 | 🔸 辅助脚本 | ✖ | 🟢 | ✖ | ✖ | 🟢 **完整 Python\u002FYAML 评估 — 灵活但较慢** |\n\n### 飞行员助手\n\n| 框架 | 开源 | GitHub |\n|--------------------------------------------|-----------------------------------------------------------------------------------------------------------------------------|------------------------------------------------------------------------------------------|\n| [Aider](https:\u002F\u002Faider.chat\u002F) | [![GitHub 仓库星级](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FAider-AI\u002Faider?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FAider-AI\u002Faider) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FAider-AI?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FAider-AI) |\n| [Cursor](https:\u002F\u002Fwww.cursor.com\u002F) | [![GitHub 仓库星级](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fgetcursor\u002Fcursor?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fgetcursor\u002Fcursor) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fgetcursor?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fgetcursor) |\n| [Continue](https:\u002F\u002Fdocs.continue.dev\u002F) | [![GitHub 仓库星级](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fcontinuedev\u002Fcontinue?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fcontinuedev\u002Fcontinue) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fcontinuedev?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fcontinuedev) |\n\n\n### 语音 API\n\n| 框架 | 开源 | GitHub |\n|--------------------------------------------|-----------------------------------------------------------------------------------------------------------------------------|------------------------------------------------------------------------------------------|\n| [VAPI.ai](https:\u002F\u002Fvapi.ai\u002F) | 🔴 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fvapiai?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fvapiai) |\n| [Bland.ai](https:\u002F\u002Fwww.bland.ai\u002F) | 🔴 | 无 |\n| [CallAnnie](https:\u002F\u002Fcallannie.ai\u002F) | 🔴 | 无 |\n| [RealtimeTTS](https:\u002F\u002Fgithub.com\u002FKoljaB\u002FRealtimeTTS) | [![GitHub 仓库星级](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FKoljaB\u002FRealtimeTTS?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FKoljaB\u002FRealtimeTTS) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FKoljaB?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FKoljaB) |\n| [RealtimeSTT](https:\u002F\u002Fgithub.com\u002FKoljaB\u002FRealtimeSTT) | [![GitHub 仓库星级](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FKoljaB\u002FRealtimeSTT?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FKoljaB\u002FRealtimeSTT) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FKoljaB?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FKoljaB) |\n| [Coqui TTS](https:\u002F\u002Fgithub.com\u002Fcoqui-ai\u002FTTS) | [![GitHub 仓库星级](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fcoqui-ai\u002FTTS?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fcoqui-ai\u002FTTS) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fcoqui-ai?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fcoqui-ai) |\n\n### 开源 TTS 模型\n\n| 模型 | 许可证 | 星级\u002F点赞数 | 过去一个月下载量 | 仓库 |\n|----------------------------------------------|------------------------------|----------------|-------------------------|----------------|\n| [Kokoro-82M](https:\u002F\u002Fhuggingface.co\u002Fhexgrad\u002FKokoro-82M) | Apache 2.0 | ⭐ 3.16k (HF) | 📥 557,392 | [Hugging Face](https:\u002F\u002Fhuggingface.co\u002Fhexgrad\u002FKokoro-82M) |\n| [Zonos-v0.1-transformer](https:\u002F\u002Fhuggingface.co\u002FZyphra\u002FZonos-v0.1-transformer) | Apache 2.0 | ⭐ 249 (HF) | 📥 24,240 | [Hugging Face](https:\u002F\u002Fhuggingface.co\u002FZyphra\u002FZonos-v0.1-transformer) |\n| [XTTS-v2](https:\u002F\u002Fhuggingface.co\u002Fcoqui\u002FXTTS-v2) | 非商业用途 | ❤️ 368 (HF) | 📥 2,545,850 | [Hugging Face](https:\u002F\u002Fhuggingface.co\u002Fcoqui\u002FXTTS-v2) |\n| [ChatTTS](https:\u002F\u002Fgithub.com\u002F2noise\u002FChatTTS) | AGPL-3.0 | 无 | 无 | [GitHub](https:\u002F\u002Fgithub.com\u002F2noise\u002FChatTTS) |\n| [MeloTTS](https:\u002F\u002Fgithub.com\u002Fmyshell-ai\u002FMeloTTS) | MIT | 无 | 无 | [GitHub](https:\u002F\u002Fgithub.com\u002Fmyshell-ai\u002FMeloTTS) |\n\n更多 TTS 模型及排名，请查看 [TTS 排行榜](https:\u002F\u002Fhuggingface.co\u002Fspaces\u002FPendrokar\u002FTTS-Spaces-Arena)。\n\n### 大模型应用框架\n\n| 工具 | 组织 | 描述 | 开源 | GitHub |\n|------|--------------|-------------|-------------|--------|\n| [Eino](https:\u002F\u002Fgithub.com\u002Fcloudwego\u002Feino) | CloudWeGo | 一个轻量级的大模型应用框架，用于构建可扩展的AI解决方案。 | [![GitHub仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fcloudwego\u002Feino?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fcloudwego\u002Feino) | [![GitHub关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fcloudwego?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fcloudwego) |\n| [对话知识挖掘解决方案加速器](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002FConversation-Knowledge-Mining-Solution-Accelerator) | 微软 | 一个利用大模型集成对话智能和知识挖掘的解决方案加速器。 | [![GitHub仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fmicrosoft\u002FConversation-Knowledge-Mining-Solution-Accelerator?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002FConversation-Knowledge-Mining-Solution-Accelerator) | [![GitHub关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fmicrosoft?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fmicrosoft) |\n| [Olmocr](https:\u002F\u002Fgithub.com\u002Fallenai\u002Folmocr) | AllenAI | 一个针对与语言模型集成优化的OCR框架。 | [![GitHub仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fallenai\u002Folmocr?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fallenai\u002Folmocr) | [![GitHub关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fallenai?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fallenai) |\n| [PDFMathTranslate](https:\u002F\u002Fgithub.com\u002FByaidu\u002FPDFMathTranslate) | Byaidu | 一个使用大模型将PDF中的数学内容转换并翻译的工具。 | [![GitHub仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FByaidu\u002FPDFMathTranslate?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FByaidu\u002FPDFMathTranslate) | [![GitHub关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FByaidu?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FByaidu) |\n| [Podcastfy](https:\u002F\u002Fgithub.com\u002Fsouzatharsis\u002Fpodcastfy) | souzatharsis | 一个利用大模型将书面内容生成播客的工具。 | [![GitHub仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fsouzatharsis\u002Fpodcastfy?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fsouzatharsis\u002Fpodcastfy) | [![GitHub关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fsouzatharsis?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fsouzatharsis) |\n| [Pandas AI](https:\u002F\u002Fgithub.com\u002Fsinaptik-ai\u002Fpandas-ai) | sinaptik-ai | 将大模型驱动的分析能力引入pandas数据框中。 | [![GitHub仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fsinaptik-ai\u002Fpandas-ai?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fsinaptik-ai\u002Fpandas-ai) | [![GitHub关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fsinaptik-ai?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fsinaptik-ai) |\n| [Ramalama](https:\u002F\u002Fgithub.com\u002Fcontainers\u002Framalama) | containers | 一个用于容器化部署AI解决方案的大模型应用框架。 | [![GitHub仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fcontainers\u002Framalama?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fcontainers\u002Framalama) | [![GitHub关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fcontainers?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fcontainers) |\n| [Robyn](https:\u002F\u002Fgithub.com\u002Ffacebookexperimental\u002FRobyn) | facebookexperimental | 来自Facebook Experimental的可扩展大模型应用框架。 | [![GitHub仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Ffacebookexperimental\u002FRobyn?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Ffacebookexperimental\u002FRobyn) | [![GitHub关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Ffacebookexperimental?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Ffacebookexperimental) |\n| [ExtractThinker](https:\u002F\u002Fgithub.com\u002Fenoch3712\u002FExtractThinker) | enoch3712 | 一个利用大模型从文本数据中提取和综合洞察的工具。 | [![GitHub仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fenoch3712\u002FExtractThinker?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fenoch3712\u002FExtractThinker) | [![GitHub关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fenoch3712?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fenoch3712) |\n\n### 操作系统 RAG 框架\n\n| 框架 | 组织 | 开源 | 发布年份 | GitHub |\n|-----------------------------------------------------|--------------|-----------------------------------------------------------------------------------------------------------------------------|----------|------------------------------------------------------------------------------------------|\n| [Haystack](https:\u002F\u002Fhaystack.deepset.ai\u002F) | deepset.ai | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fdeepset-ai\u002Fhaystack?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fdeepset-ai\u002Fhaystack) | 2023 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fdeepset-ai?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fdeepset-ai) |\n| [RAGflow](https:\u002F\u002Fragflow.io\u002F) | Infiniflow | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Finfiniflow\u002Fragflow?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Finfiniflow\u002Fragflow) | 2024 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Finfiniflow?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Finfiniflow) |\n| [txtai](https:\u002F\u002Fneuml.github.io\u002Ftxtai\u002F) | Neuml | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fneuml\u002Ftxtai?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fneuml\u002Ftxtai) | 2022 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fneuml?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fneuml) |\n| [LLM App](https:\u002F\u002Fpathway.com\u002Fdevelopers\u002Ftemplates) | Pathway | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fpathwaycom\u002Fllm-app?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fpathwaycom\u002Fllm-app) | 2023 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fpathwaycom?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fpathwaycom) |\n| [Cognita](https:\u002F\u002Fcognita.truefoundry.com\u002F) | Truefoundry | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Ftruefoundry\u002Fcognita?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Ftruefoundry\u002Fcognita) | 2024 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Ftruefoundry?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Ftruefoundry) |\n| [R2R](https:\u002F\u002Fr2r-docs.sciphi.ai\u002Fintroduction) | SciPhi AI | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FSciPhi-AI\u002FR2R?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FSciPhi-AI\u002FR2R) | 2024 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FSciPhi-AI?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FSciPhi-AI) |\n| [Raptor](https:\u002F\u002Farxiv.org\u002Fabs\u002F2401.18059) | Parth Sarthi | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fparthsarthi03\u002Fraptor?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fparthsarthi03\u002Fraptor) | 2024 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fparthsarthi03?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fparthsarthi03) |\n| [LightRAG](https:\u002F\u002Fgithub.com\u002FHKUDS\u002FLightRAG) | HKUDS | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FHKUDS\u002FLightRAG?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FHKUDS\u002FLightRAG) | 2023 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FHKUDS?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FHKUDS) |\n| [PIKE-RAG](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002FPIKE-RAG) | Microsoft | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fmicrosoft\u002FPIKE-RAG?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002FPIKE-RAG) | 2024 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fmicrosoft?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fmicrosoft) |\n| [KAG](https:\u002F\u002Fgithub.com\u002FOpenSPG\u002FKAG) | OpenSPG | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FOpenSPG\u002FKAG?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FOpenSPG\u002FKAG) | 2024 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FOpenSPG?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FOpenSPG) |\n| [MemoRAG](https:\u002F\u002Fgithub.com\u002Fqhjqhj00\u002FMemoRAG) | qhjqhj00 | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fqhjqhj00\u002FMemoRAG?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fqhjqhj00\u002FMemoRAG) | 2023 | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fqhjqhj00?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fqhjqhj00) |\n\n如果遇到困难，可以参考 [RAG_Techniques](https:\u002F\u002Fgithub.com\u002FNirDiamant\u002FRAG_Techniques)（不一定总是需要）。\n\n### 🔍 向量数据库 – 开源、性能、定价、开发体验\n\n| 向量数据库 | 许可证 | ⚡ 性能 \u002F 吞吐量 | ⏱️ 延迟（实际场景） | ☁️ 云服务定价 \u002F 免费层 | 💻 开发体验 |\n|------------------|----------------|-------------------------------------------------------------------------------------|--------------------------------------------------------------|-------------------------------------------------------------------------------|----------------------------------------------------------------------------------------------------|\n| **Qdrant** | Apache 2.0 | 🥇 单节点基准测试中 RPS 最高、延迟最低（较上一轮提升 ≥4 倍） | p95 \u003C 10ms，处理 100 万向量时（单线程） | 始终在线 1GB 免费；按需付费约 $0.014\u002F小时 | REST + gRPC；7 种语言客户端；支持过滤感知的 HNSW 索引；混合检索支持；Python“嵌入式”模式 |\n| **Milvus \u002F Zilliz Cloud** | Apache 2.0 | 🚀 索引构建速度最快；但在高维向量场景下，RPS 略逊于 Qdrant | p95 ≈ 10–20ms，处理 100 万条 768 维数据时（DiskANN，厂商数据） | 5GB 免费；无服务器方案 $0.30\u002FGB-月；专用集群从 $99\u002F月起 | 新 SDK v2（异步、Schema 缓存）；支持 Python\u002FGo\u002FJava\u002FNode |\n| **Weaviate** | BSD-3-Clause | ⚙️ 基准测试性能提升最小，但召回率尚可（95%+）且吞吐量不错 | 宣称“低毫秒级”延迟；用户反馈若配置不当，延迟可达 100–300ms | 起价 $25\u002F月；提供 14 天免费沙盒 | GraphQL + REST；强大的 SDK（Py\u002FTS\u002FGo\u002FJava）；易于实现 RAG 和混合检索模板 |\n| **pgvector** | MIT | 🔥 在 99% 召回率下，p95 比 Pinecone s1 低 28 倍，QPS 高 16 倍（5000 万条记录） | p95 \u003C 50ms，处理 5000 万条 768 维数据时（Timescale 测试） | Neon\u002FSupabase 提供免费 Postgres 数据库，并内置 pgvector（0.5–1GB，约 200 小时 CPU） | 纯 SQL；支持 JOIN 和 ACID 事务；非常适合文本与密集向量的混合查询 |\n| **Redis 8 Vector** | AGPLv3 \u002F RSAL \u002F SSPL | 🧵 在召回率 ≥0.98 的情况下，QPS 比 Qdrant 高 3.4 倍，比 Milvus 高 4 倍 | 平均亚毫秒级，负载下 \u003C10ms（厂商数据）；比 Aurora+pgvector 低 9.7 倍 | Redis Cloud：30MB 免费，按需付费从 $5\u002F月起；Flex 方案 $0.007\u002F小时 | Redis Vector Library + RAG 辅助工具；OM 客户端支持 .NET\u002FPy\u002FJS；快速部署 |\n\n### 💾 向量数据库云服务定价（2000 字分块，约 768 维）\n\n| 分块数量 | 数据大小 | 🟣 Milvus \u002F Zilliz Cloud（无服务器） | 🟢 Qdrant Cloud | 🟡 Weaviate Cloud（“标准”套餐） |\n|----------|-------------|------------------------------------------------------|-------------------------------------------------|---------------------------------------------------------------|\n| **1 万** | ~0.07 GB | 🆓 免费——在 5GB 免费层级内 | 🆓 免费——可容纳于 1GB RAM 和 4GB 磁盘中 | $25 基础费用 + $1.2 按维度收费 ≈ **$26** |\n| **10 万** | ~0.67 GB | 🆓 仍处于 5GB 以内 | 🆓 通过压缩后可容纳于 4GB 磁盘中 | $25 + $12.0 按维度收费 ≈ **$37** |\n| **100 万** | ~6.7 GB | 💵 存储费用约 $2；另加 vCU 费用或选择 $99 的专用集群 | 💵 需要 10GB 集群 → ≈ **$20\u002F月** | $25 + $120.5 按维度收费 ≈ **$145** |\n| **1000 万** | ~67 GB | 💵 存储费用约 $20；加上计算成本：**总计 $100–150** | 💵 需要 64GB 以上 → 估算为 **$120–150\u002F月** | $25 + $1,204 按维度收费 ≈ **$1,230** |\n\n### 🧠 嵌入生成成本——OpenAI（小型模型，按分块大小）\n\n| 分块数量 | 📏 1,000 字 | 📏 2,000 字 | 📏 3,000 字 |\n|--------------|---------------:|---------------:|---------------:|\n| **1,000** | $0.01 | $0.01 | $0.01 |\n| **10,000** | $0.05 | $0.10 | $0.15 |\n| **100,000** | $0.50 | $1.00 | $1.50 |\n| **1,000,000** | $5.00 | $10.00 | $15.00 |\n| **10,000,000**| $50.00 | $100.00 | $150.00 |\n\n### AI 工具（用于“使用”）\n\n| 工具 | 组织 | 描述 | 开源 | GitHub |\n|------|--------------|-------------|-------------|--------|\n| [magic-resume](https:\u002F\u002Fgithub.com\u002FJOYCEQL\u002Fmagic-resume) | JOYCEQL | 一款基于 AI 的简历生成工具。 | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FJOYCEQL\u002Fmagic-resume?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FJOYCEQL\u002Fmagic-resume) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FJOYCEQL?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FJOYCEQL) |\n| [VideoCaptioner](https:\u002F\u002Fgithub.com\u002FWEIFENG2333\u002FVideoCaptioner) | WEIFENG2333 | 一款用于自动生成视频字幕的 AI 工具。 | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FWEIFENG2333\u002FVideoCaptioner?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FWEIFENG2333\u002FVideoCaptioner) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FWEIFENG2333?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FWEIFENG2333) |\n| [DeepSeekAI](https:\u002F\u002Fgithub.com\u002FDeepLifeStudio\u002FDeepSeekAI) | DeepLifeStudio | 用于调用 DeepSeek AI 大模型的浏览器扩展。 | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FDeepLifeStudio\u002FDeepSeekAI?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FDeepLifeStudio\u002FDeepSeekAI) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FDeepLifeStudio?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FDeepLifeStudio) |\n| [logocreator](https:\u002F\u002Fgithub.com\u002FNutlope\u002Flogocreator) | Nutlope | 一款利用 AI 创作 logo 的工具。 | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FNutlope\u002Flogocreator?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FNutlope\u002Flogocreator) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FNutlope?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FNutlope) |\n| [blinkshot](https:\u002F\u002Fgithub.com\u002FNutlope\u002Fblinkshot) | Nutlope | 一款基于 AI 的截图捕捉与增强工具。 | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FNutlope\u002Fblinkshot?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FNutlope\u002Fblinkshot) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FNutlope?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FNutlope) |\n| [pollinations](https:\u002F\u002Fgithub.com\u002Fpollinations\u002Fpollinations) | pollinations | 一款利用 AI 生成创意图像和艺术作品的工具。 | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fpollinations\u002Fpollinations?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fpollinations\u002Fpollinations) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fpollinations?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fpollinations) |\n| [PromptWizard](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002FPromptWizard) | microsoft | 一款用于生成、管理和优化 AI 模型提示词的工具。 | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fmicrosoft\u002FPromptWizard?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002FPromptWizard) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fmicrosoft?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fmicrosoft) |\n| [Open-Interface](https:\u002F\u002Fgithub.com\u002FAmberSahdev\u002FOpen-Interface) | AmberSahdev | 使用 LLM 控制任何计算机。 | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FAmberSahdev\u002FOpen-Interface?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FAmberSahdev\u002FOpen-Interface) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FAmberSahdev?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FAmberSahdev) |\n| [wut](https:\u002F\u002Fgithub.com\u002Fshobrook\u002Fwut) | shobrook | 适用于终端的 LLM。 | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fshobrook\u002Fwut?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fshobrook\u002Fwut) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fshobrook?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fshobrook) |\n\n\n### 训练\u002F优化\n\n| 工具 | 组织 | 描述 | 开源 | GitHub |\n|------|--------------|-------------|-------------|--------|\n| [transformerlab-app](https:\u002F\u002Fgithub.com\u002Ftransformerlab\u002Ftransformerlab-app) | transformerlab | 一款用于训练和优化 Transformer 模型的应用程序。 | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Ftransformerlab\u002Ftransformerlab-app?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Ftransformerlab\u002Ftransformerlab-app) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Ftransformerlab?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Ftransformerlab) |\n| [fluxgym](https:\u002F\u002Fgithub.com\u002Fcocktailpeanut\u002Ffluxgym) | cocktailpeanut | 一个用于强化学习训练和优化的 Gym 环境。 | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fcocktailpeanut\u002Ffluxgym?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fcocktailpeanut\u002Ffluxgym) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fcocktailpeanut?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fcocktailpeanut) |\n| [AutoGPTQ](https:\u002F\u002Fgithub.com\u002FAutoGPTQ\u002FAutoGPTQ) | AutoGPTQ | 一款用于自动化 GPT 量化和优化的工具。 | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FAutoGPTQ\u002FAutoGPTQ?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FAutoGPTQ\u002FAutoGPTQ) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FAutoGPTQ?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FAutoGPTQ) |\n\n### AI 模型\n\n| 工具 | 组织 | 描述 | 开源 | GitHub |\n|------|--------------|-------------|-------------|--------|\n| [WALDO](https:\u002F\u002Fgithub.com\u002Fstephansturges\u002FWALDO) | stephansturges | 一个用于视觉推理和目标检测的AI模型。 | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fstephansturges\u002FWALDO?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fstephansturges\u002FWALDO) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fstephansturges?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fstephansturges) |\n| [Janus](https:\u002F\u002Fgithub.com\u002Fdeepseek-ai\u002FJanus) | deepseek-ai | 一个用于高级数据处理的多模态AI模型。 | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fdeepseek-ai\u002FJanus?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fdeepseek-ai\u002FJanus) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fdeepseek-ai?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fdeepseek-ai) |\n| [ModernBERT](https:\u002F\u002Fgithub.com\u002FAnswerDotAI\u002FModernBERT) | AnswerDotAI | BERT 的现代化版本，适用于自然语言处理任务。 | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FAnswerDotAI\u002FModernBERT?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FAnswerDotAI\u002FModernBERT) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FAnswerDotAI?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FAnswerDotAI) |\n| [Magma](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002FMagma) | microsoft | 一个用于大规模数据分析的可扩展AI模型。 | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fmicrosoft\u002FMagma?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002FMagma) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fmicrosoft?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fmicrosoft) |\n| [Cosmos-Nemotron](https:\u002F\u002Fgithub.com\u002FNVlabs\u002FCosmos-Nemotron) | NVlabs | 一个用于高级图像和视频处理的AI模型。 | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FNVlabs\u002FCosmos-Nemotron?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FNVlabs\u002FCosmos-Nemotron) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FNVlabs?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FNVlabs) |\n| [Paints-UNDO](https:\u002F\u002Fgithub.com\u002Flllyasviel\u002FPaints-UNDO) | lllyasviel | 一个用于图像生成和编辑的交互式AI模型。 | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Flllyasviel\u002FPaints-UNDO?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Flllyasviel\u002FPaints-UNDO) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Flllyasviel?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Flllyasviel) |\n\n\n### 监控\n\n| 工具 | 组织 | 描述 | 开源 | GitHub |\n|------|--------------|-------------|-------------|--------|\n| [helicone](https:\u002F\u002Fgithub.com\u002FHelicone\u002Fhelicone) | Helicone | 一个用于监控和分析AI模型性能的平台。 | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FHelicone\u002Fhelicone?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FHelicone\u002Fhelicone) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FHelicone?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FHelicone) |\n| [langwatch](https:\u002F\u002Fgithub.com\u002Flangwatch\u002Flangwatch) | langwatch | 一个用于监控语言模型输出和性能的工具。 | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Flangwatch\u002Flangwatch?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Flangwatch\u002Flangwatch) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Flangwatch?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Flangwatch) |\n\n\n### 基础设施\n\n| 工具 | 组织 | 描述 | 开源 | GitHub |\n|------|--------------|-------------|-------------|--------|\n| [gpustack](https:\u002F\u002Fgithub.com\u002Fgpustack\u002Fgpustack) | gpustack | 一个用于管理AI工作负载GPU基础设施的工具包。 | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fgpustack\u002Fgpustack?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fgpustack\u002Fgpustack) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fgpustack?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fgpustack) |\n| [harbor](https:\u002F\u002Fgithub.com\u002Fav\u002Fharbor) | av | 一个用于容器化AI基础设施管理的仓库。 | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fav\u002Fharbor?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fav\u002Fharbor) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fav?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fav) |\n\n### 思维链提示相关研究论文\n\n| 发表日期 | 标题 | 🔗 | 作者 | 机构 | 技术方法 |\n|----------------|--------------------------------------------------|--------------|--------------------|--------------------------|--------------------------|\n| 2022年1月28日 | 思维链提示激发大型语言模型的推理能力 | [🔗](https:\u002F\u002Farxiv.org\u002Fabs\u002F2201.11903) | Jason Wei 等 | DeepMind | 思维链提示 |\n| 2022年3月21日 | 自洽性提升语言模型的思维链推理能力 | [🔗](https:\u002F\u002Farxiv.org\u002Fabs\u002F2203.11171) | Xuezhi Wang 等 | DeepMind | 带自洽性的思维链提示 |\n| 2022年5月21日 | 由简入繁提示使大型语言模型具备复杂推理能力 | [🔗](https:\u002F\u002Farxiv.org\u002Fabs\u002F2205.10625) | Denny Zhou 等 | DeepMind | 由简入繁提示 |\n| 2022年5月21日 | 大型语言模型是零样本推理者 | [🔗](https:\u002F\u002Farxiv.org\u002Fabs\u002F2205.11916) | Takeshi Kojima 等 | DeepMind | 零样本思维链提示 |\n| 2022年10月6日 | ReAct：在语言模型中协同推理与行动 | [🔗](https:\u002F\u002Farxiv.org\u002Fabs\u002F2210.03629) | Shunyu Yao 等 | 普林斯顿大学 | ReAct |\n| 2023年4月1日 | 教导大型语言模型进行自我调试 | [🔗](https:\u002F\u002Farxiv.org\u002Fabs\u002F2304.05128) | Xiang Lisa Li 等 | DeepMind、斯坦福大学 | 自我调试 |\n| 2023年5月6日 | 计划-求解提示：提升大型语言模型的零样本思维链推理能力 | [🔗](https:\u002F\u002Farxiv.org\u002Fabs\u002F2305.04091) | Lei Wang 等 | 香港中文大学、商汤科技研究院 | 计划-求解提示 |\n| 2023年5月23日 | 让我们逐步验证 | [🔗](https:\u002F\u002Farxiv.org\u002Fpdf\u002F2305.20050) | Anya Goyal 等 | DeepMind | 思维链验证 |\n| 2023年10月3日 | 大型语言模型目前仍无法自我纠正推理错误 | [🔗](https:\u002F\u002Farxiv.org\u002Fabs\u002F2310.01798) | Qingxiu Dong 等 | 香港中文大学、华为诺亚方舟实验室 | LLM中的自我纠正 |\n| 2023年11月 | 大型语言模型生成的通用自洽性 | [🔗](https:\u002F\u002Farxiv.org\u002Fpdf\u002F2311.17311) | Xinyun Chen、Renat Aksitov、Uri Alon、Jie Ren、Kefan Xiao、Pengcheng Yin、Sushant Prakash、Charles Sutton、Xuezhi Wang、Denny Zhou | DeepMind | 通用自洽性 |\n| 2023年5月17日 | 思维之树：利用大型语言模型进行深思熟虑的问题解决 | [🔗](https:\u002F\u002Fexport.arxiv.org\u002Fabs\u002F2305.10601) | Shunyu Yao 等 | 普林斯顿大学、DeepMind | 思维之树 |\n| 2024年2月15日 | 无需提示的思维链推理 | [🔗](https:\u002F\u002Farxiv.org\u002Fabs\u002F2402.10200) | Xuezhi Wang、Denny Zhou | DeepMind | 思维链解码 |\n| 2024年3月21日 | ChainLM：通过改进思维链提示赋能大型语言模型 | [🔗](https:\u002F\u002Farxiv.org\u002Fabs\u002F2403.14238) | Xiaoxue Cheng 等 | 中国人民大学 | CoTGenius |\n| 2024年6月 | 语言智能体树搜索统一了语言模型中的推理、行动和规划 | [🔗](https:\u002F\u002Farxiv.org\u002Fpdf\u002F2310.04406) | Andy Zhou、Kai Yan、Michal Shlapentokh-Rothman、Haohan Wang、Yu-Xiong Wang | 无 | 语言智能体树搜索 (LATS) |\n| 2024年5月 | 蒙特卡洛树搜索通过迭代偏好学习提升推理能力 | [🔗](https:\u002F\u002Farxiv.org\u002Fpdf\u002F2405.00451) | Yuxi Xie 等 | 新加坡国立大学、DeepMind | MCTS |\n| 2024年9月18日 | 要思维链还是不要？思维链主要有助于数学和符号推理 | [🔗](https:\u002F\u002Farxiv.org\u002Fabs\u002F2409.12183) | Zayne Sprague 等 | 德克萨斯大学奥斯汀分校、约翰霍普金斯大学、普林斯顿大学 | 思维链的元分析 |\n| 2024年9月25日 | 无思维链？对规划中思维链的分析 | [🔗](https:\u002F\u002Farxiv.org\u002Fabs\u002F2305.12147) | Kaya Stechly 等 | 亚利桑那州立大学 | 规划中思维链的分析 |\n| 2024年10月18日 | 监督式思维链 | [🔗](https:\u002F\u002Farxiv.org\u002Fabs\u002F2410.14198) | Xiang Zhang、Dujian Ding | 不列颠哥伦比亚大学 | 监督式思维链 |\n| 2024年10月24日 | 关于示例：对思维链的理论理解——连贯推理与误差感知示范 | [🔗](https:\u002F\u002Farxiv.org\u002Fabs\u002F2410.16540) | Zhiqiang Hu 等 | 亚马逊、密歇根州立大学 | 思维链的理论分析 |\n\n### 思考链实现\n\n| 实现 | 链接 | 作者 | GitHub 星数 | GitHub 关注者 |\n|----------------|------|---------|--------------|------------------|\n| CoT | [chain-of-thought-hub](https:\u002F\u002Fgithub.com\u002FFranxYao\u002Fchain-of-thought-hub) | Franx Yao | [![Stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FFranxYao\u002Fchain-of-thought-hub?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FFranxYao\u002Fchain-of-thought-hub) | [![Followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FFranxYao?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FFranxYao) |\n| CoT | [optillm](https:\u002F\u002Fgithub.com\u002Fcodelion\u002Foptillm) | Codelion | [![Stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fcodelion\u002Foptillm?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fcodelion\u002Foptillm) | [![Followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fcodelion?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fcodelion) |\n| CoT | [auto-cot](https:\u002F\u002Fgithub.com\u002Famazon-science\u002Fauto-cot) | Amazon Science | [![Stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Famazon-science\u002Fauto-cot?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Famazon-science\u002Fauto-cot) | [![Followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Famazon-science?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Famazon-science) |\n| CoT | [g1](https:\u002F\u002Fgithub.com\u002Fbklieger-groq\u002Fg1) | BKlieger Groq | [![Stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fbklieger-groq\u002Fg1?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fbklieger-groq\u002Fg1) | [![Followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fbklieger-groq?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fbklieger-groq) |\n| 解码式CoT | [optillm\u002Fcot_decoding.py](https:\u002F\u002Fgithub.com\u002Fcodelion\u002Foptillm\u002Fblob\u002Fmain\u002Foptillm\u002Fcot_decoding.py) | Codelion | [![Stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fcodelion\u002Foptillm?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fcodelion\u002Foptillm) | [![Followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fcodelion?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fcodelion) |\n| 思维树 | [tree-of-thought-llm](https:\u002F\u002Fgithub.com\u002Fprinceton-nlp\u002Ftree-of-thought-llm) | Princeton NLP | [![Stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fprinceton-nlp\u002Ftree-of-thought-llm?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fprinceton-nlp\u002Ftree-of-thought-llm) | [![Followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fprinceton-nlp?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fprinceton-nlp) |\n| 思维树 | [tree-of-thoughts](https:\u002F\u002Fgithub.com\u002Fkyegomez\u002Ftree-of-thoughts) | Kye Gomez | [![Stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fkyegomez\u002Ftree-of-thoughts?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fkyegomez\u002Ftree-of-thoughts) | [![Followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fkyegomez?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fkyegomez) |\n| 思维树 | [saplings](https:\u002F\u002Fgithub.com\u002Fshobrook\u002Fsaplings) | Shobrook | [![Stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fshobrook\u002Fsaplings?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fshobrook\u002Fsaplings) | [![Followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fshobrook?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fshobrook) |\n| MCTS | [optillm\u002Fmcts.py](https:\u002F\u002Fgithub.com\u002Fcodelion\u002Foptillm\u002Fblob\u002Fmain\u002Foptillm\u002Fmcts.py) | Codelion | [![Stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fcodelion\u002Foptillm?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fcodelion\u002Foptillm) | [![Followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fcodelion?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fcodelion) |\n| 思维图 | [graph-of-thoughts](https:\u002F\u002Fgithub.com\u002Fspcl\u002Fgraph-of-thoughts) | SPCL | [![Stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fspcl\u002Fgraph-of-thoughts?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fspcl\u002Fgraph-of-thoughts) | [![Followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fspcl?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fspcl) |\n| 其他 | [CPO](https:\u002F\u002Fgithub.com\u002Fsail-sg\u002FCPO) | SAIL SG | [![Stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fsail-sg\u002FCPO?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fsail-sg\u002FCPO) | [![Followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fsail-sg?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fsail-sg) |\n| 其他 | [Everything-of-Thoughts-XoT](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002FEverything-of-Thoughts-XoT) | Microsoft | [![Stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fmicrosoft\u002FEverything-of-Thoughts-XoT?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002FEverything-of-Thoughts-XoT) | [![Followers](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fmicrosoft?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fmicrosoft) |\n\n### 思考链微调模型与数据集\n\n#### 模型\n| 模型名称 | 作者 | 参数量 | 链接 |\n|------------|--------|------|------|\n| CoT-T5-3B | KAIST AI | 30亿 | [🔗](https:\u002F\u002Fhuggingface.co\u002Fkaist-ai\u002FCoT-T5-3B) |\n| CoT-T5-11B | KAIST AI | 110亿 | [🔗](https:\u002F\u002Fhuggingface.co\u002Fkaist-ai\u002FCoT-T5-11B) |\n| Llama-3.2V-11B-cot | Xkev | 110亿 | [🔗](https:\u002F\u002Fhuggingface.co\u002FXkev\u002FLlama-3.2V-11B-cot) |\n| Llama-3.1-8B-Instruct-Reasoner-1o1_v0.3 | Lyte | 80亿 | [🔗](https:\u002F\u002Fhuggingface.co\u002FLyte\u002FLlama-3.1-8B-Instruct-Reasoner-1o1_v0.3) |\n\n\n#### 数据集\n| 数据集名称 | 作者 | 数据规模 | 点赞数 | 链接 |\n|--------------|--------|-----------|-------|------|\n| chain-of-thought-sharegpt | Isaiah Bjork | 7,140条 | 🌟 8 | [🔗](https:\u002F\u002Fhuggingface.co\u002Fdatasets\u002Fisaiahbjork\u002Fchain-of-thought-sharegpt) |\n| CoT-Collection | KAIST AI | 184万条 | 🌟 122 | [🔗](https:\u002F\u002Fhuggingface.co\u002Fdatasets\u002Fkaist-ai\u002FCoT-Collection?row=4) |\n| Reasoner-1o1-v0.3-HQ | Lyte | 370条 | 🌟 7 | [🔗](https:\u002F\u002Fhuggingface.co\u002Fdatasets\u002FLyte\u002FReasoner-1o1-v0.3-HQ) |\n| OpenLongCoT-Pretrain | qq8933 | 103,000条 | 🌟 86 | [🔗](https:\u002F\u002Fhuggingface.co\u002Fdatasets\u002Fqq8933\u002FOpenLongCoT-Pretrain?row=0) |\n\n### 学习资源\n\n| 工具 | 组织 | 描述 | 开源 | GitHub |\n|------|--------------|-------------|-------------|--------|\n| [awesome-cursorrules](https:\u002F\u002Fgithub.com\u002FPatrickJS\u002Fawesome-cursorrules) | PatrickJS | 一份关于 cursorrules 的精选资源和指南列表。 | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FPatrickJS\u002Fawesome-cursorrules?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FPatrickJS\u002Fawesome-cursorrules) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FPatrickJS?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FPatrickJS) |\n| [ai-engineering-hub](https:\u002F\u002Fgithub.com\u002Fpatchy631\u002Fai-engineering-hub) | patchy631 | 一个汇集了 AI 工程学习资源、教程和最佳实践的中心。 | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fpatchy631\u002Fai-engineering-hub?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fpatchy631\u002Fai-engineering-hub) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fpatchy631?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fpatchy631) |\n| [GenAI_Agents](https:\u002F\u002Fgithub.com\u002FNirDiamant\u002FGenAI_Agents) | NirDiamant | 构建生成式 AI 代理的相关资源和示例。 | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FNirDiamant\u002FGenAI_Agents?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FNirDiamant\u002FGenAI_Agents) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FNirDiamant?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FNirDiamant) |\n| [learn-agentic-ai](https:\u002F\u002Fgithub.com\u002Fpanaversity\u002Flearn-agentic-ai) | panaversity | 用于理解和构建自主智能体 AI 的学习材料。 | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fpanaversity\u002Flearn-agentic-ai?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fpanaversity\u002Flearn-agentic-ai) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fpanaversity?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fpanaversity) |\n| [awesome-generative-ai](https:\u002F\u002Fgithub.com\u002Fsteven2358\u002Fawesome-generative-ai) | steven2358 | 一份精选的生成式 AI 资源和项目列表。 | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fsteven2358\u002Fawesome-generative-ai?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fsteven2358\u002Fawesome-generative-ai) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fsteven2358?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fsteven2358) |\n| [awesome-mcp-servers](https:\u002F\u002Fgithub.com\u002Fpunkpeye\u002Fawesome-mcp-servers) | punkpeye | 一份精选的 MCP 服务器资源合集。 | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fpunkpeye\u002Fawesome-mcp-servers?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fpunkpeye\u002Fawesome-mcp-servers) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fpunkpeye?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fpunkpeye) |\n| [GenAI-Showcase](https:\u002F\u002Fgithub.com\u002Fmongodb-developer\u002FGenAI-Showcase) | mongodb-developer | 一个展示创新生成式 AI 项目的平台。 | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fmongodb-developer\u002FGenAI-Showcase?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fmongodb-developer\u002FGenAI-Showcase) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fmongodb-developer?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fmongodb-developer) |\n| [well-architected-iac-analyzer](https:\u002F\u002Fgithub.com\u002Faws-samples\u002Fwell-architected-iac-analyzer) | aws-samples | 一个用于分析并确保良好架构的基础设施即代码实践的工具。 | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Faws-samples\u002Fwell-architected-iac-analyzer?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Faws-samples\u002Fwell-architected-iac-analyzer) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Faws-samples?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Faws-samples) |\n| [llama-cookbook](https:\u002F\u002Fgithub.com\u002Fmeta-llama\u002Fllama-cookbook) | meta-llama | 一套用于操作 LLaMA 模型的配方和指南。 | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fmeta-llama\u002Fllama-cookbook?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fmeta-llama\u002Fllama-cookbook) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fmeta-llama?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fmeta-llama) |\n| [optillm](https:\u002F\u002Fgithub.com\u002Fcodelion\u002Foptillm) | codelion | 有关优化 LLM 使用和性能的资源。 | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fcodelion\u002Foptillm?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fcodelion\u002Foptillm) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fcodelion?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fcodelion) |\n| [cursor.directory](https:\u002F\u002Fgithub.com\u002Fpontusab\u002Fcursor.directory) | pontusab | 一个与基于光标的 workflows 相关的工具和资源目录。 | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fpontusab\u002Fcursor.directory?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002Fpontusab\u002Fcursor.directory) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002Fpontusab?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002Fpontusab) |\n| [GenAI_Agents](https:\u002F\u002Fgithub.com\u002FNirDiamant\u002FGenAI_Agents) | NirDiamant | 一份精选的生成式 AI 代理及相关工具合集。 | [![GitHub 仓库星标](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FNirDiamant\u002FGenAI_Agents?style=flat-square&color=purple)](https:\u002F\u002Fgithub.com\u002FNirDiamant\u002FGenAI_Agents) | [![GitHub 关注者](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Ffollowers\u002FNirDiamant?style=flat-square&color=teal)](https:\u002F\u002Fgithub.com\u002FNirDiamant) |","# Awesome-LLM-Resources-List 快速上手指南\n\n`Awesome-LLM-Resources-List` 并非单一的可安装软件，而是一个精选的大语言模型（LLM）资源集合，涵盖了模型托管、推理框架、本地运行工具及 API 服务平台。本指南将指导开发者如何根据需求选择并快速启动其中最流行的**本地推理工具**和**服务框架**。\n\n## 环境准备\n\n在开始使用列表中的工具前，请确保满足以下基础环境要求：\n\n* **操作系统**: Linux (推荐 Ubuntu 20.04+), macOS, 或 Windows (WSL2)。\n* **硬件要求**:\n * **CPU 运行**: 至少 8GB 内存（推荐 16GB+），适用于小参数模型量化版本。\n * **GPU 运行**: NVIDIA GPU (显存建议 8GB+)，需安装正确的 CUDA 驱动。\n* **前置依赖**:\n * Python 3.8+\n * Git\n * (可选) Docker & Docker Compose\n * (可选) NVIDIA Container Toolkit (用于 GPU 容器化部署)\n\n> **国内加速建议**:\n> * **Python 包**: 使用清华或阿里镜像源安装依赖。\n> ```bash\n> pip install -i https:\u002F\u002Fpypi.tuna.tsinghua.edu.cn\u002Fsimple \u003Cpackage_name>\n> ```\n> * **Hugging Face 模型**: 设置镜像环境变量以加速模型下载。\n> ```bash\n> export HF_ENDPOINT=https:\u002F\u002Fhf-mirror.com\n> ```\n> * **GitHub**: 若克隆仓库速度慢，建议使用国内镜像站或配置代理。\n\n---\n\n## 安装步骤\n\n根据您的需求场景，选择以下一种主流工具进行安装：\n\n### 方案 A：本地快速体验 (推荐新手) - Ollama\n最适合在本地笔记本电脑或个人服务器上快速运行开源模型。\n\n```bash\n# Linux\u002FmacOS 一键安装\ncurl -fsSL https:\u002F\u002Follama.com\u002Finstall.sh | sh\n\n# Windows 用户请下载安装包运行: https:\u002F\u002Follama.com\u002Fdownload\u002Fwindows\n\n# 验证安装\nollama --version\n```\n\n### 方案 B：高性能服务部署 - vLLM\n适合生产环境，提供高吞吐量的模型推理服务。\n\n```bash\n# 创建虚拟环境 (推荐)\npython -m venv vllm-env\nsource vllm-env\u002Fbin\u002Factivate # Windows: vllm-env\\Scripts\\activate\n\n# 安装 vLLM (NVIDIA GPU 环境)\npip install -i https:\u002F\u002Fpypi.tuna.tsinghua.edu.cn\u002Fsimple vllm\n\n# 或者使用 Docker 运行 (推荐，避免环境冲突)\ndocker run --gpus all -it --rm \\\n -p 8000:8000 \\\n --ipc=host \\\n vllm\u002Fvllm-openai:latest\n```\n\n### 方案 C：无服务器托管探索 - Modal\n适合希望将模型部署到云端且按量付费的开发者。\n\n```bash\n# 安装 Modal CLI\npip install -i https:\u002F\u002Fpypi.tuna.tsinghua.edu.cn\u002Fsimple modal\n\n# 认证账号 (按提示操作)\nmodal token new\n\n# 初始化项目\nmodal init\n```\n\n---\n\n## 基本使用\n\n### 1. 使用 Ollama 运行模型\n最简单的单行命令即可下载并运行模型（例如 Llama 3）。\n\n```bash\n# 拉取并运行模型\nollama run llama3\n\n# 在交互式命令行中直接对话\n>>> Hello, who are you?\n```\n\n*后台服务模式 (作为 API 使用):*\n```bash\n# 启动服务 (默认端口 11434)\nollama serve\n\n# 另开终端测试 API\ncurl http:\u002F\u002Flocalhost:11434\u002Fapi\u002Fgenerate -d '{\n \"model\": \"llama3\",\n \"prompt\": \"Hello!\"\n}'\n```\n\n### 2. 使用 vLLM 部署 OpenAI 兼容服务\n启动一个高性能推理服务器，完全兼容 OpenAI API 格式。\n\n```bash\n# 启动服务 (自动下载模型)\npython -m vllm.entrypoints.openai.api_server \\\n --model meta-llama\u002FMeta-Llama-3-8B-Instruct \\\n --port 8000\n\n# 调用示例 (使用 curl)\ncurl http:\u002F\u002Flocalhost:8000\u002Fv1\u002Fchat\u002Fcompletions \\\n -H \"Content-Type: application\u002Fjson\" \\\n -d '{\n \"model\": \"meta-llama\u002FMeta-Llama-3-8B-Instruct\",\n \"messages\": [{\"role\": \"user\", \"content\": \"你好，请介绍一下你自己\"}]\n }'\n```\n\n### 3. 使用 Modal 部署无服务器模型\n创建一个简单的 `main.py` 文件，利用 Modal 的云端算力。\n\n```python\n# main.py\nimport modal\n\napp = modal.App(\"example-llama\")\nimage = modal.Image.debian_slim().pip_install(\"torch\", \"transformers\")\n\n@app.function(image=image, gpu=\"A10G\")\ndef generate(prompt: str) -> str:\n from transformers import pipeline\n generator = pipeline(\"text-generation\", model=\"TinyLlama\u002FTinyLlama-1.1B-Chat-v1.0\")\n return generator(prompt)[0][\"generated_text\"]\n\nif __name__ == \"__main__\":\n # 本地触发云端运行\n print(generate.remote(\"Once upon a time,\"))\n```\n\n**运行命令:**\n```bash\nmodal run main.py\n```\n\n### 4. 访问现成 API (无需部署)\n如果您不想自行托管，可直接调用列表中提供的 API 服务（如 Together.ai, Groq 等）。\n\n```python\n# 示例：使用 OpenAI 兼容客户端调用第三方服务\nfrom openai import OpenAI\n\nclient = OpenAI(\n base_url=\"https:\u002F\u002Fapi.together.xyz\u002Fv1\", # 替换为具体服务商地址\n api_key=\"YOUR_API_KEY\"\n)\n\nresponse = client.chat.completions.create(\n model=\"meta-llama\u002FLlama-3-8b-chat-hf\",\n messages=[{\"role\": \"user\", \"content\": \"Hello world\"}]\n)\n\nprint(response.choices[0].message.content)\n```","一家初创公司的后端团队急需将开源大模型（如 Llama 3）以低成本、无服务器的方式部署到生产环境，以支持其智能客服功能。\n\n### 没有 Awesome-LLM-Resources-List 时\n- **选型迷茫**：面对 Baseten、Modal、Replicate 等十几个平台，团队花费数天查阅零散文档，仍无法判断哪个支持“一键部署”或具备真正的免费额度。\n- **成本失控风险**：由于缺乏清晰的计费对比（如 GB-sec 单价与并发限制），误选了冷启动慢且按分钟计费的方案，导致低流量时段也在白白烧钱。\n- **开发效率低下**：反复尝试不同平台的配置脚本，因缺少官方指引或社区验证的 Helper 脚本，在环境适配和 API 对接上浪费了大量工时。\n- **技术决策盲目**：不清楚哪些平台真正支持开源模型私有化部署，容易陷入供应商锁定或遭遇隐性功能限制。\n\n### 使用 Awesome-LLM-Resources-List 后\n- **快速精准选型**：直接通过表格对比\"Scale Down\"时间和\"One-Click\"支持情况，迅速锁定 Modal 或 RunPod 等适合突发流量的秒级弹性平台。\n- **成本透明可控**：参考详细的\"Serverless Compute Pricing\"数据，结合免费额度（如 AWS Lambda 的 100 万次请求）制定最优预算，避免资源浪费。\n- **部署极速落地**：利用列表中提供的官方指南链接和 GitHub Helper 脚本，将模型部署时间从几天缩短至几小时，甚至实现一键上线。\n- **架构决策有据**：清晰识别各平台对开源模型的支持程度及开发者体验评分，确保技术栈既灵活又符合长期维护需求。\n\nAwesome-LLM-Resources-List 将碎片化的部署信息转化为结构化决策依据，让团队从“盲目试错”转向“精准落地”，显著降低大模型工程化的门槛与成本。","https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Filsilfverskiold_Awesome-LLM-Resources-List_8509bee8.png","ilsilfverskiold","Ida","https:\u002F\u002Foss.gittoolsai.com\u002Favatars\u002Filsilfverskiold_eb069611.png",null,"Stockholm","ilsilfverskiold@gmail.com","https:\u002F\u002Fgithub.com\u002Filsilfverskiold",[80],{"name":81,"color":82,"percentage":83},"Python","#3572A5",100,514,85,"2026-04-17T04:58:43",1,"","未说明",{"notes":91,"python":89,"dependencies":92},"该仓库（Awesome-LLM-Resources-List）本身是一个资源列表，而非单一的可执行软件工具，因此没有统一的运行环境需求。它列出了多种不同的平台（如 Baseten, Modal, AWS Lambda）、API 服务（如 Groq, Replicate）和本地推理框架（如 Llama.cpp, Ollama, vLLM）。具体的操作系统、GPU、内存、Python 版本及依赖库要求取决于用户选择使用的具体工具或框架。例如，本地运行 vLLM 通常需要 Linux 环境和 NVIDIA GPU，而使用云端 API 则无本地硬件要求。",[],[44,14],[95,96,97],"large-language-models","llm","resources","2026-03-27T02:49:30.150509","2026-04-20T10:35:20.948617",[],[]]