[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"project-74056":3},{"id":4,"name":5,"fullName":6,"owner":7,"repo":5,"description":8,"homepage":9,"htmlUrl":10,"language":10,"languages":10,"totalLinesOfCode":10,"stars":11,"forks":12,"watchers":13,"openIssues":14,"contributorsCount":15,"subscribersCount":15,"size":15,"stars1d":16,"stars7d":17,"stars30d":18,"stars90d":15,"forks30d":15,"starsTrendScore":19,"compositeScore":20,"rankGlobal":10,"rankLanguage":10,"license":21,"archived":22,"fork":22,"defaultBranch":23,"hasWiki":22,"hasPages":24,"topics":25,"createdAt":10,"pushedAt":10,"updatedAt":38,"readmeContent":39,"aiSummary":40,"trendingCount":15,"starSnapshotCount":15,"syncStatus":41,"lastSyncTime":42,"discoverSource":43},74056,"awesome-LLM-resources","WangRongsheng\u002Fawesome-LLM-resources","WangRongsheng","🧑‍🚀 全世界最好的LLM资料总结（多模态生成、Agent、辅助编程、AI审稿、数据处理、模型训练、模型推理、o1 模型、MCP、小语言模型、视觉语言模型） | Summary of the world's best LLM resources. ","",null,8512,890,88,1,0,22,58,246,66,114.85,"Apache License 2.0",false,"main",true,[26,27,28,29,30,31,32,33,34,35,36,37],"awesome-list","book","course","large-language-models","llama","llm","mistral","openai","qwen","rag","retrieval-augmented-generation","webui","2026-06-12 04:01:13","![](.\u002Fassets\u002Flogo6.png)\n\n\u003Cp align=\"center\">全世界最好的大语言模型资源汇总 持续更新\u003C\u002Fp>\n\n\u003Cp align=\"center\">\n  \u003Ca href=\"https:\u002F\u002Fgithub.com\u002FWangRongsheng\u002Fawesome-LLM-resourses\">\u003Cimg src=https:\u002F\u002Fcdn.rawgit.com\u002Fsindresorhus\u002Fawesome\u002Fd7305f38d29fed78fa85652e3a63e154dd8e8829\u002Fmedia\u002Fbadge.svg >\u003C\u002Fa>\n  \u003Ca href=\"https:\u002F\u002Fgithub.com\u002FWangRongsheng\u002Fawesome-LLM-resourses\">\u003Cimg src=https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fforks\u002FWangRongsheng\u002Fawesome-LLM-resourses.svg?style=social >\u003C\u002Fa>\n  \u003Ca href=\"https:\u002F\u002Fgithub.com\u002FWangRongsheng\u002Fawesome-LLM-resourses\">\u003Cimg src=https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FWangRongsheng\u002Fawesome-LLM-resourses.svg?style=social >\u003C\u002Fa>\n  \u003Ca href=\"https:\u002F\u002Fgithub.com\u002FWangRongsheng\u002Fawesome-LLM-resourses\">\u003Cimg src=https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fwatchers\u002FWangRongsheng\u002Fawesome-LLM-resourses.svg?style=social >\u003C\u002Fa>\n  \u003Ca href=\"https:\u002F\u002Fgitcode.com\u002Fwangrongsheng\u002Fawesome-LLM-resources\">\u003Cimg src=\"https:\u002F\u002Fraw.githubusercontent.com\u002FWangRongsheng\u002Fawesome-LLM-resources\u002Fmain\u002Fassets\u002Fgitcode.png\" height=\"25\" alt=\"gitcode\">\n\u003C\u002Fa>\n\u003C\u002Fp>\n\n> [!TIP]\n> 如果您对**医疗数据集\u002F大模型\u002F多模态\u002F评估相关资源感兴趣**！请访问我们的 🤗 [Awesome-AI4Med](https:\u002F\u002Fgithub.com\u002FFreedomIntelligence\u002FAwesome-AI4Med) !\n> \n> 如果您希望赞助此项目，欢迎邮件联系：**wrs6@88.com**\n> \n> 赞助项目会被置顶显示在该仓库！\n\n---\n\n#### Contents\n\n- [推荐 Suggestion](#推荐-Suggestion) 🌟\n- [数据 Data](#数据-Data)\n- [微调 Fine-Tuning](#微调-Fine-Tuning) 🌟\n- [Agentic RL](#Agentic-RL) 🌟\n- [推理 Inference](#推理-Inference)\n- [评估 Evaluation](#评估-Evaluation)\n- [体验 Usage](#体验-Usage)\n- [知识库 RAG](#知识库-RAG)\n- [智能体 Agents](#智能体-Agents)\n- [研究 Research](#研究-Research) 🔥\n- [代码 Coding](#代码-Coding)\n- [视频 Video](#视频-Video) 🌟\n- [图片 Image](#图片-Image) 🌟\n- [搜索 Search](#搜索-Search)\n- [语音 Speech](#语音-Speech) 🌟\n- [龙虾 OpenClaw](#龙虾-OpenClaw) 🔥\n- [统一模型 Unified Model](#统一模型-Unified-Model) 🌟\n- [书籍 Book](#书籍-Book)\n- [课程 Course](#课程-Course)\n- [教程 Tutorial](#教程-Tutorial)\n- [论文 Paper](#论文-Paper)\n- [社区 Community](#社区-Community)\n- [模型上下文协议 MCP](#模型上下文协议-MCP)\n- [技能 Skills](#技能-Skills) 🔥\n- [推理 Open o1](#推理-Open-o1)\n- [推理 Open o3](#推理-Open-o3)\n- [小语言模型 Small Language Model](#小语言模型-Small-Language-Model) 🌟\n- [小多模态模型 Small Vision Language Model](#小多模态模型-Small-Vision-Language-Model) 🌟\n- [技巧 Tips](#技巧-tips)\n\n![](https:\u002F\u002Fcamo.githubusercontent.com\u002F2722992d519a722218f896d5f5231d49f337aaff4514e78bd59ac935334e916a\u002F68747470733a2f2f692e696d6775722e636f6d2f77617856496d762e706e67)\n\n## 推荐 Suggestion\n\n#### Podcast\n\n- [140. 对姚顺宇的4小时访谈：请允许我小疯一下！在Anthropic和Gemini训模型、技术预测、英雄主义已过去](https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=Gk_KUg3qED0)\n- [张驰: A Year Inside ByteDance's AI Lab](https:\u002F\u002Fchangche.substack.com\u002Fp\u002Fa-year-inside-bytedances-ai-lab)\n- [Luo Fuli: OpenClaw, Agent Frameworks — The AI Paradigm Has Already Changed Dramatically!](https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=V9eI-t3TApE)\n- [A 7-hour marathon interview with Saining Xie: World Models, AMI Labs, Yann LeCun, Fei-Fei Li, and 42](https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=rIwgZWzUKm8)\n- [翁家翌：OpenAI，GPT，强化学习，Infra，后训练，天授，tuixue，开源，CMU，清华｜WhynotTV Podcast](https:\u002F\u002Fwww.bilibili.com\u002Fvideo\u002FBV1darmBcE4A?vd_source=c739db1ebdd361d47af5a0b8497417db)\n- [Lovart 创始人陈冕×罗永浩！且让我大闹一场，然后悄然离去](https:\u002F\u002Fwww.bilibili.com\u002Fvideo\u002FBV14eiQBmEbN\u002F?spm_id_from=333.1387.upload.video_card.click)\n- [MiniMax 创始人闫俊杰×罗永浩！大山并非无法翻越](https:\u002F\u002Fwww.bilibili.com\u002Fvideo\u002FBV11NmtBzE36\u002F?spm_id_from=333.1387.upload.video_card.click&vd_source=c739db1ebdd361d47af5a0b8497417db)\n- [影视飓风TIM×罗永浩！用影像打开世界的梦想家](https:\u002F\u002Fwww.bilibili.com\u002Fvideo\u002FBV1B5xkzPEhx\u002F?spm_id_from=333.1387.upload.video_card.click&vd_source=c739db1ebdd361d47af5a0b8497417db)\n- [129. 全球大模型第一股的上市访谈，和智谱CEO张鹏聊：敢问路在何方？](https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=9zSMTUUEfmU&list=PLwAchVoh-4zNSI5UlKEkKCL5r_jJyrFeO&index=2)\n- [128. Manus决定出售前最后的访谈：啊，这奇幻的2025年漂流啊…](https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=MW-ezf2RhVg&list=PLwAchVoh-4zNSI5UlKEkKCL5r_jJyrFeO&index=3)\n- [122. 朱啸虎现实主义故事的第三次连载：人工智能的盛筵与泡泡](https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=wK0-m3rKgZ0&list=PLwAchVoh-4zNSI5UlKEkKCL5r_jJyrFeO&index=9)\n- [119. Kimi Linear、Minimax M2？和杨松琳考古算法变种史，并预演未来架构改进方案](https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=858HR43pegk&list=PLwAchVoh-4zNSI5UlKEkKCL5r_jJyrFeO&index=12&t=1070s)\n- [118. 对李想的第二次3小时访谈：CEO大模型、MoE、梁文锋、VLA、能量、记忆、对抗人性、亲密关系、人类的智慧](https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=RxXVq7-sJzM&list=PLwAchVoh-4zNSI5UlKEkKCL5r_jJyrFeO&index=13)\n- [115. 对OpenAI姚顺雨3小时访谈：6年Agent研究、人与系统、吞噬的边界、既单极又多元的世界](https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=gQgKkUsx5q0&list=PLwAchVoh-4zNSI5UlKEkKCL5r_jJyrFeO&index=16)\n- [113. 和杨植麟时隔1年的对话：K2、Agentic LLM、缸中之脑和“站在无限的开端”](https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=ouG6jrkECrc&list=PLwAchVoh-4zNSI5UlKEkKCL5r_jJyrFeO&index=18)\n\n## 数据 Data\n\n> [!NOTE]\n> \n> 此处命名为`数据`，但这里并没有提供具体数据集，而是提供了处理获取大规模数据的方法\n\n\n1. [AotoLabel](https:\u002F\u002Fgithub.com\u002Frefuel-ai\u002Fautolabel): Label, clean and enrich text datasets with LLMs.\n2. [LabelLLM](https:\u002F\u002Fgithub.com\u002Fopendatalab\u002FLabelLLM): The Open-Source Data Annotation Platform.\n3. [data-juicer](https:\u002F\u002Fgithub.com\u002Fmodelscope\u002Fdata-juicer): A one-stop data processing system to make data higher-quality, juicier, and more digestible for LLMs!\n4. [OmniParser](https:\u002F\u002Fgithub.com\u002Fjf-tech\u002Fomniparser): a native Golang ETL streaming parser and transform library for CSV, JSON, XML, EDI, text, etc.\n5. [MinerU (`🔥`)](https:\u002F\u002Fgithub.com\u002Fopendatalab\u002FMinerU): MinerU is a one-stop, open-source, high-quality data extraction tool, supports PDF\u002Fwebpage\u002Fe-book extraction.\n6. [PDF-Extract-Kit](https:\u002F\u002Fgithub.com\u002Fopendatalab\u002FPDF-Extract-Kit): A Comprehensive Toolkit for High-Quality PDF Content Extraction.\n7. [Parsera](https:\u002F\u002Fgithub.com\u002Fraznem\u002Fparsera): Lightweight library for scraping web-sites with LLMs.\n8. [Sparrow](https:\u002F\u002Fgithub.com\u002Fkatanaml\u002Fsparrow): Sparrow is an innovative open-source solution for efficient data extraction and processing from various documents and images.\n9. [Docling](https:\u002F\u002Fgithub.com\u002FDS4SD\u002Fdocling): Get your documents ready for gen AI.\n10. [GOT-OCR2.0](https:\u002F\u002Fgithub.com\u002FUcas-HaoranWei\u002FGOT-OCR2.0): OCR Model.\n11. [LLM Decontaminator](https:\u002F\u002Fgithub.com\u002Flm-sys\u002Fllm-decontaminator): Rethinking Benchmark and Contamination for Language Models with Rephrased Samples.\n12. [DataTrove](https:\u002F\u002Fgithub.com\u002Fhuggingface\u002Fdatatrove): DataTrove is a library to process, filter and deduplicate text data at a very large scale.\n13. [llm-swarm](https:\u002F\u002Fgithub.com\u002Fhuggingface\u002Fllm-swarm\u002Ftree\u002Fmain\u002Fexamples\u002Ftextbooks): Generate large synthetic datasets like [Cosmopedia](https:\u002F\u002Fhuggingface.co\u002Fdatasets\u002FHuggingFaceTB\u002Fcosmopedia).\n14. [Distilabel](https:\u002F\u002Fgithub.com\u002Fargilla-io\u002Fdistilabel): Distilabel is a framework for synthetic data and AI feedback for engineers who need fast, reliable and scalable pipelines based on verified research papers.\n15. [Common-Crawl-Pipeline-Creator](https:\u002F\u002Fhuggingface.co\u002Fspaces\u002Flhoestq\u002FCommon-Crawl-Pipeline-Creator): The Common Crawl Pipeline Creator.\n16. [Tabled](https:\u002F\u002Fgithub.com\u002FVikParuchuri\u002Ftabled): Detect and extract tables to markdown and csv.\n17. [Zerox](https:\u002F\u002Fgithub.com\u002Fgetomni-ai\u002Fzerox): Zero shot pdf OCR with gpt-4o-mini.\n18. [DocLayout-YOLO](https:\u002F\u002Fgithub.com\u002Fopendatalab\u002FDocLayout-YOLO): Enhancing Document Layout Analysis through Diverse Synthetic Data and Global-to-Local Adaptive Perception.\n19. [TensorZero](https:\u002F\u002Fgithub.com\u002Ftensorzero\u002Ftensorzero): make LLMs improve through experience.\n20. [Promptwright](https:\u002F\u002Fgithub.com\u002FStacklokLabs\u002Fpromptwright): Generate large synthetic data using a local LLM.\n21. [pdf-extract-api](https:\u002F\u002Fgithub.com\u002FCatchTheTornado\u002Fpdf-extract-api): Document (PDF) extraction and parse API using state of the art modern OCRs + Ollama supported models.\n22. [pdf2htmlEX](https:\u002F\u002Fgithub.com\u002Fpdf2htmlEX\u002Fpdf2htmlEX): Convert PDF to HTML without losing text or format.\n23. [Extractous](https:\u002F\u002Fgithub.com\u002Fyobix-ai\u002Fextractous): Fast and efficient unstructured data extraction. Written in Rust with bindings for many languages.\n24. [MegaParse](https:\u002F\u002Fgithub.com\u002FQuivrHQ\u002FMegaParse): File Parser optimised for LLM Ingestion with no loss.\n25. [MarkItDown](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fmarkitdown): Python tool for converting files and office documents to Markdown.\n26. [datasketch](https:\u002F\u002Fgithub.com\u002Fekzhu\u002Fdatasketch): datasketch gives you probabilistic data structures that can process and search very large amount of data super fast, with little loss of accuracy.\n27. [semhash](https:\u002F\u002Fgithub.com\u002FMinishLab\u002Fsemhash): lightweight and flexible tool for deduplicating datasets using semantic similarity.\n28. [ReaderLM-v2](https:\u002F\u002Fhuggingface.co\u002Fjinaai\u002FReaderLM-v2): a 1.5B parameter language model that converts raw HTML into beautifully formatted markdown or JSON.\n29. [Bespoke Curator](https:\u002F\u002Fgithub.com\u002Fbespokelabsai\u002Fcurator): Data Curation for Post-Training & Structured Data Extraction.\n30. [LangKit](https:\u002F\u002Fgithub.com\u002Fwhylabs\u002Flangkit): An open-source toolkit for monitoring Large Language Models (LLMs). Extracts signals from prompts & responses, ensuring safety & security.\n31. [Curator](https:\u002F\u002Fgithub.com\u002Fbespokelabsai\u002Fcurator): Synthetic Data curation for post-training and structured data extraction.\n32. [olmOCR](https:\u002F\u002Fgithub.com\u002Fallenai\u002Folmocr): A toolkit for training language models to work with PDF documents in the wild.\n33. [Easy Dataset (`🔥`)](https:\u002F\u002Fgithub.com\u002FConardLi\u002Feasy-dataset): A powerful tool for creating fine-tuning datasets for LLM.\n34. [BabelDOC](https:\u002F\u002Fgithub.com\u002Ffunstory-ai\u002FBabelDOC): PDF scientific paper translation and bilingual comparison library.\n35. [Dolphin](https:\u002F\u002Fgithub.com\u002Fbytedance\u002FDolphin): Document Image Parsing via Heterogeneous Anchor Prompting.\n36. [EasyDistill](https:\u002F\u002Fgithub.com\u002Fmodelscope\u002Feasydistill): Easy Knowledge Distillation for Large Language Models.\n37. [ContextGem](https:\u002F\u002Fgithub.com\u002Fshcherbak-ai\u002Fcontextgem): a free, open-source LLM framework that makes it radically easier to extract structured data and insights from documents.\n38. [OCRFlux](https:\u002F\u002Fgithub.com\u002Fchatdoc-com\u002FOCRFlux): a lightweight yet powerful multimodal toolkit that significantly advances PDF-to-Markdown conversion, excelling in complex layout handling, complicated table parsing and cross-page content merging.\n39. [DataFlow](https:\u002F\u002Fgithub.com\u002FOpenDCAI\u002FDataFlow): Easy Data Preparation with latest LLMs-based Operators and Pipelines.\n40. [DatasetLoom (`multimodal`)](https:\u002F\u002Fgithub.com\u002F599yongyang\u002FDatasetLoom): 一个面向多模态大模型训练的智能数据集构建与评估平台.\n41. [Logics-Parsing](https:\u002F\u002Fgithub.com\u002Falibaba\u002FLogics-Parsing)\n42. [DeepSeek-OCR](https:\u002F\u002Fhuggingface.co\u002Fdeepseek-ai\u002FDeepSeek-OCR)\n43. [PaddleOCR-VL](https:\u002F\u002Fhuggingface.co\u002FPaddlePaddle\u002FPaddleOCR-VL)\n44. [Chandra](https:\u002F\u002Fgithub.com\u002Fdatalab-to\u002Fchandra): a highly accurate OCR model that converts images and PDFs into structured HTML\u002FMarkdown\u002FJSON while preserving layout information.\n45. [HunyuanOCR](https:\u002F\u002Fgithub.com\u002FTencent-Hunyuan\u002FHunyuanOCR): a leading end-to-end OCR expert VLM powered by Hunyuan's native multimodal architecture.\n46. [DeepSeek-OCR-2](https:\u002F\u002Fhuggingface.co\u002Fdeepseek-ai\u002FDeepSeek-OCR-2): Visual Causal Flow.\n47. [PaddleOCR-VL-1.5 (`🔥`)](https:\u002F\u002Fhuggingface.co\u002FPaddlePaddle\u002FPaddleOCR-VL-1.5): Towards a Multi-Task 0.9B VLM for Robust In-the-Wild Document Parsing.\n48. [GLM-OCR](https:\u002F\u002Fhuggingface.co\u002Fzai-org\u002FGLM-OCR): a multimodal OCR model for complex document understanding, built on the GLM-V encoder–decoder architecture.\n\n\u003Cdiv align=\"right\">\n    \u003Cb>\u003Ca href=\"#Contents\">↥ back to top\u003C\u002Fa>\u003C\u002Fb>\n\u003C\u002Fdiv>\n\n## 微调 Fine-Tuning\n\n1. [LLaMA-Factory (`🔥`)](https:\u002F\u002Fgithub.com\u002Fhiyouga\u002FLLaMA-Factory): Unify Efficient Fine-Tuning of 100+ LLMs.\n2. [360-LLaMA-Factory](https:\u002F\u002Fgithub.com\u002FQihoo360\u002F360-LLaMA-Factory): Unify Efficient Fine-Tuning of 100+ LLMs. (add Sequence Parallelism for supporting long context training)\n4. [unsloth (`🔥`)](https:\u002F\u002Fgithub.com\u002Funslothai\u002Funsloth): 2-5X faster 80% less memory LLM finetuning.\n5. [TRL](https:\u002F\u002Fhuggingface.co\u002Fdocs\u002Ftrl\u002Findex): Transformer Reinforcement Learning.\n6. [Firefly](https:\u002F\u002Fgithub.com\u002Fyangjianxin1\u002FFirefly): Firefly: 大模型训练工具，支持训练数十种大模型\n7. [Xtuner](https:\u002F\u002Fgithub.com\u002FInternLM\u002Fxtuner): An efficient, flexible and full-featured toolkit for fine-tuning large models.\n8. [torchtune](https:\u002F\u002Fgithub.com\u002Fpytorch\u002Ftorchtune): A Native-PyTorch Library for LLM Fine-tuning.\n9. [Swift](https:\u002F\u002Fgithub.com\u002Fmodelscope\u002Fswift): Use PEFT or Full-parameter to finetune 200+ LLMs or 15+ MLLMs.\n10. [AutoTrain](https:\u002F\u002Fhuggingface.co\u002Fautotrain): A new way to automatically train, evaluate and deploy state-of-the-art Machine Learning models.\n11. [OpenRLHF](https:\u002F\u002Fgithub.com\u002FOpenLLMAI\u002FOpenRLHF): An Easy-to-use, Scalable and High-performance RLHF Framework (Support 70B+ full tuning & LoRA & Mixtral & KTO).\n12. [Ludwig](https:\u002F\u002Fgithub.com\u002Fludwig-ai\u002Fludwig): Low-code framework for building custom LLMs, neural networks, and other AI models.\n13. [mistral-finetune](https:\u002F\u002Fgithub.com\u002Fmistralai\u002Fmistral-finetune): A light-weight codebase that enables memory-efficient and performant finetuning of Mistral's models.\n14. [aikit](https:\u002F\u002Fgithub.com\u002Fsozercan\u002Faikit): Fine-tune, build, and deploy open-source LLMs easily!\n15. [H2O-LLMStudio](https:\u002F\u002Fgithub.com\u002Fh2oai\u002Fh2o-llmstudio): H2O LLM Studio - a framework and no-code GUI for fine-tuning LLMs.\n16. [LitGPT](https:\u002F\u002Fgithub.com\u002FLightning-AI\u002Flitgpt): Pretrain, finetune, deploy 20+ LLMs on your own data. Uses state-of-the-art techniques: flash attention, FSDP, 4-bit, LoRA, and more.\n17. [LLMBox](https:\u002F\u002Fgithub.com\u002FRUCAIBox\u002FLLMBox): A comprehensive library for implementing LLMs, including a unified training pipeline and comprehensive model evaluation.\n18. [PaddleNLP](https:\u002F\u002Fgithub.com\u002FPaddlePaddle\u002FPaddleNLP): Easy-to-use and powerful NLP and LLM library.\n19. [workbench-llamafactory](https:\u002F\u002Fgithub.com\u002FNVIDIA\u002Fworkbench-llamafactory): This is an NVIDIA AI Workbench example project that demonstrates an end-to-end model development workflow using Llamafactory.\n20. [OpenRLHF](https:\u002F\u002Fgithub.com\u002FOpenLLMAI\u002FOpenRLHF): An Easy-to-use, Scalable and High-performance RLHF Framework (70B+ PPO Full Tuning & Iterative DPO & LoRA & Mixtral).\n21. [TinyLLaVA Factory](https:\u002F\u002Fgithub.com\u002FTinyLLaVA\u002FTinyLLaVA_Factory): A Framework of Small-scale Large Multimodal Models.\n22. [LLM-Foundry](https:\u002F\u002Fgithub.com\u002Fmosaicml\u002Fllm-foundry): LLM training code for Databricks foundation models.\n23. [lmms-finetune](https:\u002F\u002Fgithub.com\u002Fzjysteven\u002Flmms-finetune): A unified codebase for finetuning (full, lora) large multimodal models, supporting llava-1.5, qwen-vl, llava-interleave, llava-next-video, phi3-v etc.\n24. [Simplifine](https:\u002F\u002Fgithub.com\u002Fsimplifine-llm\u002FSimplifine): Simplifine lets you invoke LLM finetuning with just one line of code using any Hugging Face dataset or model.\n25. [Transformer Lab](https:\u002F\u002Fgithub.com\u002Ftransformerlab\u002Ftransformerlab-app): Open Source Application for Advanced LLM Engineering: interact, train, fine-tune, and evaluate large language models on your own computer.\n26. [Liger-Kernel](https:\u002F\u002Fgithub.com\u002Flinkedin\u002FLiger-Kernel): Efficient Triton Kernels for LLM Training.\n27. [ChatLearn](https:\u002F\u002Fgithub.com\u002Falibaba\u002FChatLearn): A flexible and efficient training framework for large-scale alignment.\n28. [nanotron](https:\u002F\u002Fgithub.com\u002Fhuggingface\u002Fnanotron): Minimalistic large language model 3D-parallelism training.\n29. [Proxy Tuning](https:\u002F\u002Fgithub.com\u002Falisawuffles\u002Fproxy-tuning): Tuning Language Models by Proxy.\n30. [Effective LLM Alignment](https:\u002F\u002Fgithub.com\u002FVikhrModels\u002Feffective_llm_alignment\u002F): Effective LLM Alignment Toolkit.\n31. [Autotrain-advanced](https:\u002F\u002Fgithub.com\u002Fhuggingface\u002Fautotrain-advanced)\n32. [Meta Lingua](https:\u002F\u002Fgithub.com\u002Ffacebookresearch\u002Flingua): a lean, efficient, and easy-to-hack codebase to research LLMs.\n33. [Vision-LLM Alignemnt](https:\u002F\u002Fgithub.com\u002FNiuTrans\u002FVision-LLM-Alignment): This repository contains the code for SFT, RLHF, and DPO, designed for vision-based LLMs, including the LLaVA models and the LLaMA-3.2-vision models.\n34. [finetune-Qwen2-VL](https:\u002F\u002Fgithub.com\u002Fzhangfaen\u002Ffinetune-Qwen2-VL): Quick Start for Fine-tuning or continue pre-train Qwen2-VL Model.\n35. [Online-RLHF](https:\u002F\u002Fgithub.com\u002FRLHFlow\u002FOnline-RLHF): A recipe for online RLHF and online iterative DPO.\n36. [InternEvo](https:\u002F\u002Fgithub.com\u002FInternLM\u002FInternEvo): an open-sourced lightweight training framework aims to support model pre-training without the need for extensive dependencies.\n37. [veRL (`🔥`)](https:\u002F\u002Fgithub.com\u002Fvolcengine\u002Fverl): Volcano Engine Reinforcement Learning for LLM.\n38. [Axolotl](https:\u002F\u002Faxolotl-ai-cloud.github.io\u002Faxolotl\u002F): Axolotl is designed to work with YAML config files that contain everything you need to preprocess a dataset, train or fine-tune a model, run model inference or evaluation, and much more.\n39. [Oumi](https:\u002F\u002Fgithub.com\u002Foumi-ai\u002Foumi): Everything you need to build state-of-the-art foundation models, end-to-end.\n40. [Kiln](https:\u002F\u002Fgithub.com\u002FKiln-AI\u002FKiln): The easiest tool for fine-tuning LLM models, synthetic data generation, and collaborating on datasets.\n41. [DeepSeek-671B-SFT-Guide](https:\u002F\u002Fgithub.com\u002FScienceOne-AI\u002FDeepSeek-671B-SFT-Guide): An open-source solution for full parameter fine-tuning of DeepSeek-V3\u002FR1 671B, including complete code and scripts from training to inference, as well as some practical experiences and conclusions.\n42. [MLX-VLM](https:\u002F\u002Fgithub.com\u002FBlaizzy\u002Fmlx-vlm): MLX-VLM is a package for inference and fine-tuning of Vision Language Models (VLMs) on your Mac using MLX.\n43. [RL-Factory](https:\u002F\u002Fgithub.com\u002FSimple-Efficient\u002FRL-Factory): Train your Agent model via our easy and efficient framework.\n44. [RM-Gallery](https:\u002F\u002Fgithub.com\u002Fmodelscope\u002FRM-Gallery): A One-Stop Reward Model Platform.\n45. [ART](https:\u002F\u002Fgithub.com\u002FOpenPipe\u002FART): rain multi-step agents for real-world tasks using GRPO. Give your agents on-the-job training.\n46. [LMMs-Engine](https:\u002F\u002Fgithub.com\u002FEvolvingLMMs-Lab\u002Flmms-engine): A simple, any-to-any modality framework for pretraining and finetuning. Lean, flexible, and built for research.\n47. [dLLM](https:\u002F\u002Fgithub.com\u002FZHZisZZ\u002Fdllm): a library that unifies the training and evaluation of diffusion language models, bringing transparency and reproducibility to the entire development pipeline. `diffusion`\n48. [Miles](https:\u002F\u002Fgithub.com\u002Fradixark\u002Fmiles): an enterprise-facing reinforcement learning framework for large-scale MoE post-training and production workloads.\n49. [Skills](https:\u002F\u002Fgithub.com\u002FNVIDIA-NeMo\u002FSkills): a collection of pipelines to improve \"skills\" of large language models (LLMs).\n50. [Twinkle](https:\u002F\u002Fgithub.com\u002Fmodelscope\u002Ftwinkle): a lightweight, client-server training framework engineered with modular, high-cohesion interfaces.\n51. [NeMo AutoModel](https:\u002F\u002Fgithub.com\u002FNVIDIA-NeMo\u002FAutomodel): Pytorch Distributed native training library for LLMs\u002FVLMs with OOTB Hugging Face support.\n\n\u003Cdiv align=\"right\">\n    \u003Cb>\u003Ca href=\"#Contents\">↥ back to top\u003C\u002Fa>\u003C\u002Fb>\n\u003C\u002Fdiv>\n\n## Agentic RL\n\n- veRL (`🔥`): https:\u002F\u002Fgithub.com\u002Fvolcengine\u002Fverl\n- AReaL: https:\u002F\u002Fgithub.com\u002FinclusionAI\u002FAReaL\n- slime (`🔥`): https:\u002F\u002Fgithub.com\u002FTHUDM\u002Fslime\n- Agent Lightning: https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fagent-lightning\n\n\u003Cdiv align=\"right\">\n    \u003Cb>\u003Ca href=\"#Contents\">↥ back to top\u003C\u002Fa>\u003C\u002Fb>\n\u003C\u002Fdiv>\n\n## 推理 Inference\n\n1. [ollama (`🔥`)](https:\u002F\u002Fgithub.com\u002Follama\u002Follama): Get up and running with Llama 3, Mistral, Gemma, and other large language models.\n2. [Open WebUI](https:\u002F\u002Fgithub.com\u002Fopen-webui\u002Fopen-webui): User-friendly WebUI for LLMs (Formerly Ollama WebUI).\n3. [Text Generation WebUI](https:\u002F\u002Fgithub.com\u002Foobabooga\u002Ftext-generation-webui): A Gradio web UI for Large Language Models. Supports transformers, GPTQ, AWQ, EXL2, llama.cpp (GGUF), Llama models.\n4. [Xinference](https:\u002F\u002Fgithub.com\u002Fxorbitsai\u002Finference): A powerful and versatile library designed to serve language, speech recognition, and multimodal models.\n5. [LangChain](https:\u002F\u002Fgithub.com\u002Flangchain-ai\u002Flangchain): Build context-aware reasoning applications.\n6. [LlamaIndex](https:\u002F\u002Fgithub.com\u002Frun-llama\u002Fllama_index): A data framework for your LLM applications.\n7. [lobe-chat](https:\u002F\u002Fgithub.com\u002Flobehub\u002Flobe-chat): an open-source, modern-design LLMs\u002FAI chat framework. Supports Multi AI Providers, Multi-Modals (Vision\u002FTTS) and plugin system.\n8. [TensorRT-LLM](https:\u002F\u002Fgithub.com\u002FNVIDIA\u002FTensorRT-LLM): TensorRT-LLM provides users with an easy-to-use Python API to define Large Language Models (LLMs) and build TensorRT engines that contain state-of-the-art optimizations to perform inference efficiently on NVIDIA GPUs.\n9. [vllm (`🔥`)](https:\u002F\u002Fgithub.com\u002Fvllm-project\u002Fvllm): A high-throughput and memory-efficient inference and serving engine for LLMs.\n10. [LlamaChat](https:\u002F\u002Fgithub.com\u002Falexrozanski\u002FLlamaChat): Chat with your favourite LLaMA models in a native macOS app.\n11. [NVIDIA ChatRTX](https:\u002F\u002Fwww.nvidia.com\u002Fen-us\u002Fai-on-rtx\u002Fchatrtx\u002F): ChatRTX is a demo app that lets you personalize a GPT large language model (LLM) connected to your own content—docs, notes, or other data.\n12. [LM Studio (`🔥`)](https:\u002F\u002Flmstudio.ai\u002F): Discover, download, and run local LLMs.\n13. [chat-with-mlx](https:\u002F\u002Fgithub.com\u002Fqnguyen3\u002Fchat-with-mlx): Chat with your data natively on Apple Silicon using MLX Framework.\n14. [LLM Pricing](https:\u002F\u002Fllmpricecheck.com\u002F): Quickly Find the Perfect Large Language Models (LLM) API for Your Budget! Use Our Free Tool for Instant Access to the Latest Prices from Top Providers.\n15. [Open Interpreter](https:\u002F\u002Fgithub.com\u002FOpenInterpreter\u002Fopen-interpreter): A natural language interface for computers.\n16. [Chat-ollama](https:\u002F\u002Fgithub.com\u002Fsugarforever\u002Fchat-ollama): An open source chatbot based on LLMs. It supports a wide range of language models, and knowledge base management.\n17. [chat-ui](https:\u002F\u002Fgithub.com\u002Fhuggingface\u002Fchat-ui): Open source codebase powering the HuggingChat app.\n18. [MemGPT](https:\u002F\u002Fgithub.com\u002Fcpacker\u002FMemGPT): Create LLM agents with long-term memory and custom tools.\n19. [koboldcpp](https:\u002F\u002Fgithub.com\u002FLostRuins\u002Fkoboldcpp): A simple one-file way to run various GGML and GGUF models with KoboldAI's UI.\n20. [LLMFarm](https:\u002F\u002Fgithub.com\u002Fguinmoon\u002FLLMFarm): llama and other large language models on iOS and MacOS offline using GGML library.\n21. [enchanted](https:\u002F\u002Fgithub.com\u002FAugustDev\u002Fenchanted): Enchanted is iOS and macOS app for chatting with private self hosted language models such as Llama2, Mistral or Vicuna using Ollama.\n22. [Flowise](https:\u002F\u002Fgithub.com\u002FFlowiseAI\u002FFlowise): Drag & drop UI to build your customized LLM flow.\n23. [Jan](https:\u002F\u002Fgithub.com\u002Fjanhq\u002Fjan): Jan is an open source alternative to ChatGPT that runs 100% offline on your computer. Multiple engine support (llama.cpp, TensorRT-LLM).\n24. [LMDeploy](https:\u002F\u002Fgithub.com\u002FInternLM\u002Flmdeploy): LMDeploy is a toolkit for compressing, deploying, and serving LLMs.\n25. [RouteLLM](https:\u002F\u002Fgithub.com\u002Flm-sys\u002FRouteLLM): A framework for serving and evaluating LLM routers - save LLM costs without compromising quality!\n26. [MInference](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002FMInference): About\nTo speed up Long-context LLMs' inference, approximate and dynamic sparse calculate the attention, which reduces inference latency by up to 10x for pre-filling on an A100 while maintaining accuracy.\n27. [Mem0](https:\u002F\u002Fgithub.com\u002Fmem0ai\u002Fmem0): The memory layer for Personalized AI.\n28. [SGLang (`🔥`)](https:\u002F\u002Fgithub.com\u002Fsgl-project\u002Fsglang): SGLang is yet another fast serving framework for large language models and vision language models.\n29. [AirLLM](https:\u002F\u002Fgithub.com\u002Flyogavin\u002Fairllm): AirLLM optimizes inference memory usage, allowing 70B large language models to run inference on a single 4GB GPU card without quantization, distillation and pruning. And you can run 405B Llama3.1 on 8GB vram now.\n30. [LLMHub](https:\u002F\u002Fgithub.com\u002Fjmather\u002Fllmhub): LLMHub is a lightweight management platform designed to streamline the operation and interaction with various language models (LLMs).\n31. [YuanChat](https:\u002F\u002Fgithub.com\u002FIEIT-Yuan\u002FYuanChat)\n32. [LiteLLM (`🔥`)](https:\u002F\u002Fgithub.com\u002FBerriAI\u002Flitellm): Call all LLM APIs using the OpenAI format [Bedrock, Huggingface, VertexAI, TogetherAI, Azure, OpenAI, Groq etc.]\n33. [GuideLLM](https:\u002F\u002Fgithub.com\u002Fneuralmagic\u002Fguidellm): GuideLLM is a powerful tool for evaluating and optimizing the deployment of large language models (LLMs).\n34. [LLM-Engines](https:\u002F\u002Fgithub.com\u002Fjdf-prog\u002FLLM-Engines): A unified inference engine for large language models (LLMs) including open-source models (VLLM, SGLang, Together) and commercial models (OpenAI, Mistral, Claude).\n35. [OARC](https:\u002F\u002Fgithub.com\u002FLeoleojames1\u002Follama_agent_roll_cage): ollama_agent_roll_cage (OARC) is a local python agent fusing ollama llm's with Coqui-TTS speech models, Keras classifiers, Llava vision, Whisper recognition, and more to create a unified chatbot agent for local, custom automation.\n36. [g1](https:\u002F\u002Fgithub.com\u002Fbklieger-groq\u002Fg1): Using Llama-3.1 70b on Groq to create o1-like reasoning chains.\n37. [MemoryScope](https:\u002F\u002Fgithub.com\u002Fmodelscope\u002FMemoryScope): MemoryScope provides LLM chatbots with powerful and flexible long-term memory capabilities, offering a framework for building such abilities.\n38. [OpenLLM](https:\u002F\u002Fgithub.com\u002Fbentoml\u002FOpenLLM): Run any open-source LLMs, such as Llama 3.1, Gemma, as OpenAI compatible API endpoint in the cloud.\n39. [Infinity](https:\u002F\u002Fgithub.com\u002Finfiniflow\u002Finfinity): The AI-native database built for LLM applications, providing incredibly fast hybrid search of dense embedding, sparse embedding, tensor and full-text.\n40. [optillm](https:\u002F\u002Fgithub.com\u002Fcodelion\u002Foptillm): an OpenAI API compatible optimizing inference proxy which implements several state-of-the-art techniques that can improve the accuracy and performance of LLMs.\n41. [LLaMA Box](https:\u002F\u002Fgithub.com\u002Fgpustack\u002Fllama-box): LLM inference server implementation based on llama.cpp.\n42. [ZhiLight](https:\u002F\u002Fgithub.com\u002Fzhihu\u002FZhiLight): A highly optimized inference acceleration engine for Llama and its variants.\n43. [DashInfer](https:\u002F\u002Fgithub.com\u002Fmodelscope\u002Fdash-infer): DashInfer is a native LLM inference engine aiming to deliver industry-leading performance atop various hardware architectures.\n44. [LocalAI](https:\u002F\u002Fgithub.com\u002Fmudler\u002FLocalAI): The free, Open Source alternative to OpenAI, Claude and others. Self-hosted and local-first. Drop-in replacement for OpenAI, running on consumer-grade hardware. No GPU required.\n45. [ktransformers](https:\u002F\u002Fgithub.com\u002Fkvcache-ai\u002Fktransformers): A Flexible Framework for Experiencing Cutting-edge LLM Inference Optimizations.\n46. [SkyPilot](https:\u002F\u002Fgithub.com\u002Fskypilot-org\u002Fskypilot): Run AI and batch jobs on any infra (Kubernetes or 14+ clouds). Get unified execution, cost savings, and high GPU availability via a simple interface.\n47. [Chitu](https:\u002F\u002Fgithub.com\u002Fthu-pacman\u002Fchitu): High-performance inference framework for large language models, focusing on efficiency, flexibility, and availability.\n48. [TokenSwift](https:\u002F\u002Fgithub.com\u002Fbigai-nlco\u002FTokenSwift): From Hours to Minutes: Lossless Acceleration of Ultra Long Sequence Generation.\n49. [Cherry Studio (`🔥`)](https:\u002F\u002Fgithub.com\u002FCherryHQ\u002Fcherry-studio): a desktop client that supports for multiple LLM providers, available on Windows, Mac and Linux.\n50. [Shimmy](https:\u002F\u002Fgithub.com\u002FMichael-A-Kuykendall\u002Fshimmy): Python-free Rust inference server — OpenAI-API compatible. GGUF + SafeTensors, hot model swap, auto-discovery, single binary.\n51. [LlamaBarn](https:\u002F\u002Fgithub.com\u002Fggml-org\u002FLlamaBarn): Run local LLMs on your Mac with a simple menu bar app.\n52. [Parallax](https:\u002F\u002Fgithub.com\u002FGradientHQ\u002Fparallax): a distributed model serving framework that lets you build your own AI cluster anywhere.\n53. [xLLM](https:\u002F\u002Fgithub.com\u002Fjd-opensource\u002Fxllm): A high-performance inference engine for LLMs, optimized for diverse AI accelerators.\n\n\n\u003Cdiv align=\"right\">\n    \u003Cb>\u003Ca href=\"#Contents\">↥ back to top\u003C\u002Fa>\u003C\u002Fb>\n\u003C\u002Fdiv>\n\n## 评估 Evaluation\n\n1. [lm-evaluation-harness](https:\u002F\u002Fgithub.com\u002FEleutherAI\u002Flm-evaluation-harness): A framework for few-shot evaluation of language models.\n2. [opencompass (`🔥`)](https:\u002F\u002Fgithub.com\u002Fopen-compass\u002Fopencompass): OpenCompass is an LLM evaluation platform, supporting a wide range of models (Llama3, Mistral, InternLM2,GPT-4,LLaMa2, Qwen,GLM, Claude, etc) over 100+ datasets.\n3. [llm-comparator](https:\u002F\u002Fgithub.com\u002FPAIR-code\u002Fllm-comparator): LLM Comparator is an interactive data visualization tool for evaluating and analyzing LLM responses side-by-side, developed.\n4. [EvalScope (`🔥`)](https:\u002F\u002Fgithub.com\u002Fmodelscope\u002Fevalscope)\n5. [Weave](https:\u002F\u002Fweave-docs.wandb.ai\u002Fguides\u002Fcore-types\u002Fevaluations): A lightweight toolkit for tracking and evaluating LLM applications.\n6. [MixEval](https:\u002F\u002Fgithub.com\u002FPsycoy\u002FMixEval\u002F): Deriving Wisdom of the Crowd from LLM Benchmark Mixtures.\n7. [Evaluation guidebook](https:\u002F\u002Fgithub.com\u002Fhuggingface\u002Fevaluation-guidebook): If you've ever wondered how to make sure an LLM performs well on your specific task, this guide is for you!\n8. [Ollama Benchmark](https:\u002F\u002Fgithub.com\u002Faidatatools\u002Follama-benchmark): LLM Benchmark for Throughput via Ollama (Local LLMs).\n9. [VLMEvalKit](https:\u002F\u002Fgithub.com\u002Fopen-compass\u002FVLMEvalKit): Open-source evaluation toolkit of large vision-language models (LVLMs), support ~100 VLMs, 40+ benchmarks.\n10. [AGI-Eval](https:\u002F\u002Fagi-eval.cn\u002Fmvp\u002Fhome)\n11. [EvalScope](https:\u002F\u002Fgithub.com\u002Fmodelscope\u002Fevalscope): A streamlined and customizable framework for efficient large model evaluation and performance benchmarking.\n12. [DeepEval](https:\u002F\u002Fgithub.com\u002Fconfident-ai\u002Fdeepeval): a simple-to-use, open-source LLM evaluation framework, for evaluating and testing large-language model systems.\n13. [Lighteval](https:\u002F\u002Fgithub.com\u002Fhuggingface\u002Flighteval): Lighteval is your all-in-one toolkit for evaluating LLMs across multiple backends.\n14. [QwQ\u002Feval](https:\u002F\u002Fgithub.com\u002FQwenLM\u002FQwQ\u002Ftree\u002Fmain\u002Feval): QwQ is the reasoning model series developed by Qwen team, Alibaba Cloud.\n15. [Evalchemy](https:\u002F\u002Fgithub.com\u002Fmlfoundations\u002Fevalchemy): A unified and easy-to-use toolkit for evaluating post-trained language models.\n16. [MathArena](https:\u002F\u002Fgithub.com\u002Feth-sri\u002Fmatharena): Evaluation of LLMs on latest math competitions.\n17. [YourBench](https:\u002F\u002Fgithub.com\u002Fhuggingface\u002Fyourbench): A Dynamic Benchmark Generation Framework.\n18. [MedEvalKit](https:\u002F\u002Fgithub.com\u002Falibaba-damo-academy\u002FMedEvalKit): A Unified Medical Evaluation Framework.\n19. [OpenJudge](https:\u002F\u002Fgithub.com\u002Fmodelscope\u002FOpenJudge): A Unified Framework for Holistic Evaluation and Quality Rewards.\n\n\u003Cdiv align=\"right\">\n    \u003Cb>\u003Ca href=\"#Contents\">↥ back to top\u003C\u002Fa>\u003C\u002Fb>\n\u003C\u002Fdiv>\n\n## 体验 Usage\n\n1. [LM Arena](https:\u002F\u002Flmarena.ai\u002Fzh)\n2. [Design Arena](https:\u002F\u002Fwww.designarena.ai\u002F)\n\n\u003Cdiv align=\"right\">\n    \u003Cb>\u003Ca href=\"#Contents\">↥ back to top\u003C\u002Fa>\u003C\u002Fb>\n\u003C\u002Fdiv>\n\n## 知识库 RAG\n\n1. [AnythingLLM](https:\u002F\u002Fgithub.com\u002FMintplex-Labs\u002Fanything-llm): The all-in-one AI app for any LLM with full RAG and AI Agent capabilites.\n2. [MaxKB](https:\u002F\u002Fgithub.com\u002F1Panel-dev\u002FMaxKB): 基于 LLM 大语言模型的知识库问答系统。开箱即用，支持快速嵌入到第三方业务系统\n3. [RAGFlow](https:\u002F\u002Fgithub.com\u002Finfiniflow\u002Fragflow): An open-source RAG (Retrieval-Augmented Generation) engine based on deep document understanding.\n4. [Dify](https:\u002F\u002Fgithub.com\u002Flanggenius\u002Fdify): An open-source LLM app development platform. Dify's intuitive interface combines AI workflow, RAG pipeline, agent capabilities, model management, observability features and more, letting you quickly go from prototype to production.\n5. [FastGPT](https:\u002F\u002Fgithub.com\u002Flabring\u002FFastGPT): A knowledge-based platform built on the LLM, offers out-of-the-box data processing and model invocation capabilities, allows for workflow orchestration through Flow visualization.\n6. [Langchain-Chatchat](https:\u002F\u002Fgithub.com\u002Fchatchat-space\u002FLangchain-Chatchat): 基于 Langchain 与 ChatGLM 等不同大语言模型的本地知识库问答\n7. [QAnything](https:\u002F\u002Fgithub.com\u002Fnetease-youdao\u002FQAnything): Question and Answer based on Anything.\n8. [Quivr](https:\u002F\u002Fgithub.com\u002FQuivrHQ\u002Fquivr): A personal productivity assistant (RAG) ⚡️🤖 Chat with your docs (PDF, CSV, ...) & apps using Langchain, GPT 3.5 \u002F 4 turbo, Private, Anthropic, VertexAI, Ollama, LLMs, Groq that you can share with users ! Local & Private alternative to OpenAI GPTs & ChatGPT powered by retrieval-augmented generation.\n9. [RAG-GPT](https:\u002F\u002Fgithub.com\u002Fopen-kf\u002Frag-gpt): RAG-GPT, leveraging LLM and RAG technology, learns from user-customized knowledge bases to provide contextually relevant answers for a wide range of queries, ensuring rapid and accurate information retrieval.\n10. [Verba](https:\u002F\u002Fgithub.com\u002Fweaviate\u002FVerba): Retrieval Augmented Generation (RAG) chatbot powered by Weaviate.\n11. [FlashRAG](https:\u002F\u002Fgithub.com\u002FRUC-NLPIR\u002FFlashRAG): A Python Toolkit for Efficient RAG Research.\n12. [GraphRAG](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fgraphrag): A modular graph-based Retrieval-Augmented Generation (RAG) system.\n13. [LightRAG](https:\u002F\u002Fgithub.com\u002FSylphAI-Inc\u002FLightRAG): LightRAG helps developers with both building and optimizing Retriever-Agent-Generator pipelines.\n14. [GraphRAG-Ollama-UI](https:\u002F\u002Fgithub.com\u002Fseverian42\u002FGraphRAG-Ollama-UI): GraphRAG using Ollama with Gradio UI and Extra Features.\n15. [nano-GraphRAG](https:\u002F\u002Fgithub.com\u002Fgusye1234\u002Fnano-graphrag): A simple, easy-to-hack GraphRAG implementation.\n16. [RAG Techniques](https:\u002F\u002Fgithub.com\u002FNirDiamant\u002FRAG_Techniques): This repository showcases various advanced techniques for Retrieval-Augmented Generation (RAG) systems. RAG systems combine information retrieval with generative models to provide accurate and contextually rich responses.\n17. [ragas](https:\u002F\u002Fgithub.com\u002Fexplodinggradients\u002Fragas): Evaluation framework for your Retrieval Augmented Generation (RAG) pipelines.\n18. [kotaemon](https:\u002F\u002Fgithub.com\u002FCinnamon\u002Fkotaemon): An open-source clean & customizable RAG UI for chatting with your documents. Built with both end users and developers in mind.\n19. [RAGapp](https:\u002F\u002Fgithub.com\u002Fragapp\u002Fragapp): The easiest way to use Agentic RAG in any enterprise.\n20. [TurboRAG](https:\u002F\u002Fgithub.com\u002FMooreThreads\u002FTurboRAG): Accelerating Retrieval-Augmented Generation with Precomputed KV Caches for Chunked Text.\n21. [LightRAG](https:\u002F\u002Fgithub.com\u002FHKUDS\u002FLightRAG): Simple and Fast Retrieval-Augmented Generation.\n22. [TEN](https:\u002F\u002Fgithub.com\u002FTEN-framework\u002Ften_framework): the Next-Gen AI-Agent Framework, the world's first truly real-time multimodal AI agent framework.\n23. [AutoRAG](https:\u002F\u002Fgithub.com\u002FMarker-Inc-Korea\u002FAutoRAG): RAG AutoML tool for automatically finding an optimal RAG pipeline for your data.\n24. [KAG](https:\u002F\u002Fgithub.com\u002FOpenSPG\u002FKAG): KAG is a knowledge-enhanced generation framework based on OpenSPG engine, which is used to build knowledge-enhanced rigorous decision-making and information retrieval knowledge services.\n25. [Fast-GraphRAG](https:\u002F\u002Fgithub.com\u002Fcirclemind-ai\u002Ffast-graphrag): RAG that intelligently adapts to your use case, data, and queries.\n26. [Tiny-GraphRAG](https:\u002F\u002Fgithub.com\u002Flimafang\u002Ftiny-graphrag)\n27. [DB-GPT GraphRAG](https:\u002F\u002Fgithub.com\u002Feosphoros-ai\u002FDB-GPT\u002Ftree\u002Fmain\u002Fdbgpt\u002Fstorage\u002Fknowledge_graph): DB-GPT GraphRAG integrates both triplet-based knowledge graphs and document structure graphs while leveraging community and document retrieval mechanisms to enhance RAG capabilities, achieving comparable performance while consuming only 50% of the tokens required by Microsoft's GraphRAG. Refer to the DB-GPT [Graph RAG User Manual](http:\u002F\u002Fdocs.dbgpt.cn\u002Fdocs\u002Fcookbook\u002Frag\u002Fgraph_rag_app_develop\u002F) for details.\n28. [Chonkie](https:\u002F\u002Fgithub.com\u002Fbhavnicksm\u002Fchonkie): The no-nonsense RAG chunking library that's lightweight, lightning-fast, and ready to CHONK your texts.\n29. [RAGLite](https:\u002F\u002Fgithub.com\u002Fsuperlinear-ai\u002Fraglite): RAGLite is a Python toolkit for Retrieval-Augmented Generation (RAG) with PostgreSQL or SQLite.\n30. [KAG](https:\u002F\u002Fgithub.com\u002FOpenSPG\u002FKAG): KAG is a logical form-guided reasoning and retrieval framework based on OpenSPG engine and LLMs.\n31. [CAG](https:\u002F\u002Fgithub.com\u002Fhhhuang\u002FCAG): CAG leverages the extended context windows of modern large language models (LLMs) by preloading all relevant resources into the model’s context and caching its runtime parameters.\n32. [MiniRAG](https:\u002F\u002Fgithub.com\u002FHKUDS\u002FMiniRAG): an extremely simple retrieval-augmented generation framework that enables small models to achieve good RAG performance through heterogeneous graph indexing and lightweight topology-enhanced retrieval.\n33. [XRAG](https:\u002F\u002Fgithub.com\u002FDocAILab\u002FXRAG): a benchmarking framework designed to evaluate the foundational components of advanced Retrieval-Augmented Generation (RAG) systems.\n34. [Rankify](https:\u002F\u002Fgithub.com\u002FDataScienceUIBK\u002Frankify): A Comprehensive Python Toolkit for Retrieval, Re-Ranking, and Retrieval-Augmented Generation.\n35. [RAG-Anything](https:\u002F\u002Fgithub.com\u002FHKUDS\u002FRAG-Anything): All-in-One RAG System.\n\n\u003Cdiv align=\"right\">\n    \u003Cb>\u003Ca href=\"#Contents\">↥ back to top\u003C\u002Fa>\u003C\u002Fb>\n\u003C\u002Fdiv>\n\n## 智能体 Agents\n\n1. [AutoGen](https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002Fautogen): AutoGen is a framework that enables the development of LLM applications using multiple agents that can converse with each other to solve tasks. [AutoGen AIStudio](https:\u002F\u002Fautogen-studio.com\u002F)\n2. [CrewAI](https:\u002F\u002Fgithub.com\u002Fjoaomdmoura\u002FcrewAI): Framework for orchestrating role-playing, autonomous AI agents. By fostering collaborative intelligence, CrewAI empowers agents to work together seamlessly, tackling complex tasks.\n3. [Coze](https:\u002F\u002Fwww.coze.com\u002F)\n4. [AgentGPT](https:\u002F\u002Fgithub.com\u002Freworkd\u002FAgentGPT): Assemble, configure, and deploy autonomous AI Agents in your browser.\n5. [XAgent](https:\u002F\u002Fgithub.com\u002FOpenBMB\u002FXAgent): An Autonomous LLM Agent for Complex Task Solving.\n6. [MobileAgent](https:\u002F\u002Fgithub.com\u002FX-PLUG\u002FMobileAgent): The Powerful Mobile Device Operation Assistant Family.\n7. [Lagent](https:\u002F\u002Fgithub.com\u002FInternLM\u002Flagent): A lightweight framework for building LLM-based agents.\n8. [Qwen-Agent](https:\u002F\u002Fgithub.com\u002FQwenLM\u002FQwen-Agent): Agent framework and applications built upon Qwen2, featuring Function Calling, Code Interpreter, RAG, and Chrome extension.\n9. [LinkAI](https:\u002F\u002Flink-ai.tech\u002Fportal): 一站式 AI 智能体搭建平台\n10. [Baidu APPBuilder](https:\u002F\u002Fappbuilder.cloud.baidu.com\u002F)\n11. [agentUniverse](https:\u002F\u002Fgithub.com\u002Falipay\u002FagentUniverse): agentUniverse is a LLM multi-agent framework that allows developers to easily build multi-agent applications. Furthermore, through the community, they can exchange and share practices of patterns across different domains.\n12. [LazyLLM](https:\u002F\u002Fgithub.com\u002FLazyAGI\u002FLazyLLM): 低代码构建多Agent大模型应用的开发工具\n13. [AgentScope](https:\u002F\u002Fgithub.com\u002Fmodelscope\u002Fagentscope): Start building LLM-empowered multi-agent applications in an easier way.\n14. [AgentField](https:\u002F\u002Fgithub.com\u002FAgent-Field\u002Fagentfield): Open-source control plane for building and operating AI agents like APIs at scale, with routing, memory, observability, identity, auth, and policy controls.\n15. [MoA](https:\u002F\u002Fgithub.com\u002Ftogethercomputer\u002FMoA): Mixture of Agents (MoA) is a novel approach that leverages the collective strengths of multiple LLMs to enhance performance, achieving state-of-the-art results.\n15. [Agently](https:\u002F\u002Fgithub.com\u002FMaplemx\u002FAgently): AI Agent Application Development Framework.\n16. [OmAgent](https:\u002F\u002Fgithub.com\u002Fom-ai-lab\u002FOmAgent): A multimodal agent framework for solving complex tasks.\n17. [Tribe](https:\u002F\u002Fgithub.com\u002FStreetLamb\u002Ftribe): No code tool to rapidly build and coordinate multi-agent teams.\n18. [CAMEL](https:\u002F\u002Fgithub.com\u002Fcamel-ai\u002Fcamel): First LLM multi-agent framework and an open-source community dedicated to finding the scaling law of agents.\n19. [PraisonAI](https:\u002F\u002Fgithub.com\u002FMervinPraison\u002FPraisonAI\u002F): PraisonAI application combines AutoGen and CrewAI or similar frameworks into a low-code solution for building and managing multi-agent LLM systems, focusing on simplicity, customisation, and efficient human-agent collaboration.\n20. [IoA](https:\u002F\u002Fgithub.com\u002Fopenbmb\u002Fioa): An open-source framework for collaborative AI agents, enabling diverse, distributed agents to team up and tackle complex tasks through internet-like connectivity.\n21. [llama-agentic-system ](https:\u002F\u002Fgithub.com\u002Fmeta-llama\u002Fllama-agentic-system): Agentic components of the Llama Stack APIs.\n22. [Agent Zero](https:\u002F\u002Fgithub.com\u002Ffrdel\u002Fagent-zero): Agent Zero is not a predefined agentic framework. It is designed to be dynamic, organically growing, and learning as you use it.\n23. [Agents](https:\u002F\u002Fgithub.com\u002Faiwaves-cn\u002Fagents): An Open-source Framework for Data-centric, Self-evolving Autonomous Language Agents.\n24. [AgentScope](https:\u002F\u002Fgithub.com\u002Fmodelscope\u002Fagentscope): Start building LLM-empowered multi-agent applications in an easier way.\n25. [FastAgency](https:\u002F\u002Fgithub.com\u002Fairtai\u002Ffastagency): The fastest way to bring multi-agent workflows to production.\n26. [Swarm](https:\u002F\u002Fgithub.com\u002Fopenai\u002Fswarm): Framework for building, orchestrating and deploying multi-agent systems. Managed by OpenAI Solutions team. Experimental framework.\n27. [Agent-S](https:\u002F\u002Fgithub.com\u002Fsimular-ai\u002FAgent-S): an open agentic framework that uses computers like a human.\n28. [PydanticAI](https:\u002F\u002Fgithub.com\u002Fpydantic\u002Fpydantic-ai): Agent Framework \u002F shim to use Pydantic with LLMs.\n29. [Agentarium](https:\u002F\u002Fgithub.com\u002FThytu\u002FAgentarium): open-source framework for creating and managing simulations populated with AI-powered agents.\n30. [smolagents](https:\u002F\u002Fgithub.com\u002Fhuggingface\u002Fsmolagents): a barebones library for agents. Agents write python code to call tools and orchestrate other agents.\n31. [Cooragent](https:\u002F\u002Fgithub.com\u002FLeapLabTHU\u002Fcooragent): Cooragent is an AI agent collaboration community.\n32. [Agno](https:\u002F\u002Fgithub.com\u002Fagno-agi\u002Fagno): Agno is a lightweight library for building Agents with memory, knowledge, tools and reasoning.\n33. [Suna](https:\u002F\u002Fgithub.com\u002Fkortix-ai\u002Fsuna): Open Source Generalist AI Agent.\n34. [rowboat](https:\u002F\u002Fgithub.com\u002Frowboatlabs\u002Frowboat): Let AI build multi-agent workflows for you in minutes.\n35. [EvoAgentX](https:\u002F\u002Fgithub.com\u002FEvoAgentX\u002FEvoAgentX): Building a Self-Evolving Ecosystem of AI Agents.\n36. [ii-agent](https:\u002F\u002Fgithub.com\u002FIntelligent-Internet\u002Fii-agent): a new open-source framework to build and deploy intelligent agents.\n37. [OWL](https:\u002F\u002Fgithub.com\u002Fcamel-ai\u002Fowl): Optimized Workforce Learning for General Multi-Agent Assistance in Real-World Task Automation.\n38. [OpenManus](https:\u002F\u002Fgithub.com\u002FFoundationAgents\u002FOpenManus): No fortress, purely open ground. OpenManus is Coming.\n39. [JoyAgent-JDGenie](https:\u002F\u002Fgithub.com\u002Fjd-opensource\u002Fjoyagent-jdgenie): 业界首个开源高完成度轻量化通用多智能体产品.\n40. [coze-studio](https:\u002F\u002Fgithub.com\u002Fcoze-dev\u002Fcoze-studio): An AI agent development platform with all-in-one visual tools, simplifying agent creation, debugging, and deployment like never before.\n41. [OxyGent](https:\u002F\u002Fgithub.com\u002Fjd-opensource\u002FOxyGent): An advanced Python framework that empowers developers to quickly build production-ready intelligent systems.\n42. [LazyCraft](https:\u002F\u002Fgithub.com\u002FLazyAGI\u002FLazyCraft): LazyCraft 是一个基于 LazyLLM 构建的 AI Agent 应用开发与管理平台，旨在协助开发者以 低门槛、低成本 快速构建和发布大模型应用。\n43. [OpenAgents](https:\u002F\u002Fgithub.com\u002Fopenagents-org\u002Fopenagents): AI Agent Networks for Open Collaboration.\n44. [SandBox](https:\u002F\u002Fgithub.com\u002Fagent-infra\u002Fsandbox): All-in-One Sandbox for AI Agents that combines Browser, Shell, File, MCP and VSCode Server in a single Docker container.\n45. [DeepAnalyze](https:\u002F\u002Fgithub.com\u002Fruc-datalab\u002FDeepAnalyze): First agentic LLM for autonomous data science, supporting specific data tasks (data preparation, analysis, modeling, visualization, and insight) and data-oriented deep research (produce analyst-grade research reports).\n46. [Astron Agent](https:\u002F\u002Fgithub.com\u002Fiflytek\u002Fastron-agent): Enterprise-grade, commercial-friendly agentic workflow platform for building next-generation SuperAgents.\n47. [Youtu-Agent](https:\u002F\u002Fgithub.com\u002FTencentCloudADP\u002Fyoutu-agent): A simple yet powerful agent framework that delivers with open-source models.\n48. [MiroThinker](https:\u002F\u002Fgithub.com\u002FMiroMindAI\u002FMiroThinker): an open-source search agent model, built for tool-augmented reasoning and real-world information seeking, aiming to match the deep research experience of OpenAI Deep Research and Gemini Deep Research.\n49. [Nexent](https:\u002F\u002Fgithub.com\u002FModelEngine-Group\u002Fnexent): A zero-code platform for auto-generating agents — no orchestration, no complex drag-and-drop required, using pure language to develop any agent you want.\n50. [Yunjue-Agent](https:\u002F\u002Fgithub.com\u002FYunjueTech\u002FYunjue-Agent): A Fully Reproducible, Zero-Start In-Situ Self-Evolving Agent System for Open-Ended Tasks.\n51. [Hindsight](https:\u002F\u002Fgithub.com\u002Fvectorize-io\u002Fhindsight): State-of-the-art long-term memory for AI agents by Vectorize. Open-source, self-hostable, with integrations for LangChain, CrewAI, LlamaIndex, MCP, and more.\n52. [AgentsMesh](https:\u002F\u002Fgithub.com\u002FAgentsMesh\u002FAgentsMesh): The AI Agent Workforce Platform. Self-hostable multi-agent orchestration with remote AI workstations (AgentPods), PTY sandbox + git worktree isolation, channels-based agent collaboration, built-in Kanban, and per-pod MCP server. Supports Claude Code, Codex CLI, Gemini CLI, Aider, OpenCode.\n\n\u003Cdiv align=\"right\">\n    \u003Cb>\u003Ca href=\"#Contents\">↥ back to top\u003C\u002Fa>\u003C\u002Fb>\n\u003C\u002Fdiv>\n\n## 研究 Research\n\n#### 写作\n\n- PaperDebugger: https:\u002F\u002Fgithub.com\u002FPaperDebugger\u002FPaperDebugger\n- Chat Overleaf: https:\u002F\u002Fgithub.com\u002Fanuin-cat\u002Fchat-overleaf\n- 文智云助手: https:\u002F\u002Foverleaf.top\u002F\n- LiteWrite: https:\u002F\u002Flitewrite.ai\u002F\n- Prism: https:\u002F\u002Fopenai.com\u002Fzh-Hans-CN\u002Fprism\u002F\n- claude-prism: https:\u002F\u002Fgithub.com\u002Fdelibae\u002Fclaude-prism\n\n#### 审稿\n\n- PaperReview: https:\u002F\u002Fpaperreview.ai\u002F\n- aiXiv: https:\u002F\u002Faixiv.science\u002F\n- OpenJudge Review: https:\u002F\u002Fopenjudge.me\u002Fpaper_review\n\n#### 其他\n\n- Paper2Video: https:\u002F\u002Fgithub.com\u002Fshowlab\u002FPaper2Video\n- Paper2Poster: https:\u002F\u002Fgithub.com\u002FPaper2Poster\u002FPaper2Poster\n- AutoPR: https:\u002F\u002Fgithub.com\u002Firgolic\u002FAutoPR\n- Auto-Slides: https:\u002F\u002Fgithub.com\u002FWestlake-AGI-Lab\u002FAuto-Slides\n- EvoPresent: https:\u002F\u002Fgithub.com\u002Feric-ai-lab\u002FEvoPresent\n- Paper2All: https:\u002F\u002Fgithub.com\u002FYuhangChen1\u002FPaper2All\n- AutoPage: https:\u002F\u002Fgithub.com\u002FAutoLab-SAI-SJTU\u002FAutoPage\n- pdf2video: https:\u002F\u002Fgithub.com\u002FDangJin\u002Fpdf2video\n- Idea2Paper: https:\u002F\u002Fgithub.com\u002FAgentAlphaAGI\u002FIdea2Paper\n- PaperX: https:\u002F\u002Fgithub.com\u002Fyutao1024\u002FPaperX\n- figures4papers: https:\u002F\u002Fgithub.com\u002FChenLiu-1996\u002Ffigures4papers\n- PaperBanana: https:\u002F\u002Fgithub.com\u002Fdwzhu-pku\u002FPaperBanana\n- PaperBanana-Pro: https:\u002F\u002Fgithub.com\u002Felpsykongloo\u002FPaperBanana-Pro\n- PPTAgent: https:\u002F\u002Fgithub.com\u002Ficip-cas\u002FPPTAgent\n- AutoFigure: https:\u002F\u002Fgithub.com\u002FResearAI\u002FAutoFigure\n  - FigureWeave: https:\u002F\u002Fgithub.com\u002FKrisocer\u002FFigureWeave\n  - EditDeck: https:\u002F\u002Fgithub.com\u002FMorgensonne\u002FEditDeck\n- AutoFigure-Edit: https:\u002F\u002Fgithub.com\u002FResearAI\u002FAutoFigure-Edit\n- Academic Figure Generator: https:\u002F\u002Fgithub.com\u002FLigphiDonk\u002Facademic-figure-generator\n- Paper PPT Agent: https:\u002F\u002Fgithub.com\u002FCRui5in\u002Fpaper-ppt-agent\n- PPT Master: https:\u002F\u002Fgithub.com\u002Fhugohe3\u002Fppt-master\n- LandPPT: https:\u002F\u002Fgithub.com\u002Fsligter\u002FLandPPT\n\n#### 全自动科研\n\n- EvoScientist: https:\u002F\u002Fgithub.com\u002FEvoScientist\u002FEvoScientist\n- Auto-claude-code-research-in-sleep: https:\u002F\u002Fgithub.com\u002Fwanshuiyin\u002FAuto-claude-code-research-in-sleep\n- ArgusBot: https:\u002F\u002Fgithub.com\u002Fwaltstephen\u002FArgusBot\n- Station: https:\u002F\u002Fgithub.com\u002Fdualverse-ai\u002Fstation\n- Dr.Claw: https:\u002F\u002Fgithub.com\u002FOpenLAIR\u002Fdr-claw\n- Redigg: https:\u002F\u002Fgithub.com\u002Fredigg\u002Fredigg\n- AutoResearchClaw: https:\u002F\u002Fgithub.com\u002Faiming-lab\u002FAutoResearchClaw\n- NanoResearch: https:\u002F\u002Fgithub.com\u002FOpenRaiser\u002FNanoResearch\n- ScienceClaw: https:\u002F\u002Fgithub.com\u002FAgentTeam-TaichuAI\u002FScienceClaw\n- EurekaClaw: https:\u002F\u002Fgithub.com\u002FEurekaClaw\u002FEurekaClaw\n- Claude-scholar: https:\u002F\u002Fgithub.com\u002FGalaxy-Dawn\u002Fclaude-scholar\u002F\n- claude-scientific-skills: https:\u002F\u002Fgithub.com\u002FK-Dense-AI\u002Fclaude-scientific-skills\n- K-Dense BYOK: https:\u002F\u002Fgithub.com\u002FK-Dense-AI\u002Fk-dense-byok\n- latex-paper-skills: https:\u002F\u002Fgithub.com\u002Fyunshenwuchuxun\u002Flatex-paper-skills\n- NeuriCo: https:\u002F\u002Fgithub.com\u002FChicagoHAI\u002FNeuriCo\n- AutoResearch : https:\u002F\u002Fgithub.com\u002Fkarpathy\u002Fautoresearch\n- RD-Agent : https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002FRD-Agent\n- DeepScientist : https:\u002F\u002Fgithub.com\u002FResearAI\u002FDeepScientist\n- Deep Researcher Agent: https:\u002F\u002Fgithub.com\u002FXiangyue-Zhang\u002Fauto-deep-researcher-24x7\n- academic-research-skills: https:\u002F\u002Fgithub.com\u002FImbad0202\u002Facademic-research-skills\n- Supervisor-Skills: https:\u002F\u002Fgithub.com\u002FHKUSTDial\u002FSupervisor-Skills\n\n\u003Cdiv align=\"right\">\n    \u003Cb>\u003Ca href=\"#Contents\">↥ back to top\u003C\u002Fa>\u003C\u002Fb>\n\u003C\u002Fdiv>\n\n## 代码 Coding\n\n1. [Cloi CLI](https:\u002F\u002Fgithub.com\u002Fcloi-ai\u002Fcloi): Local debugging agent that runs in your terminal.\n2. [Devin](https:\u002F\u002Fdevin.ai\u002F)\n3. [v0](https:\u002F\u002Fv0.dev\u002F)\n4. [Blot.new](https:\u002F\u002Fbolt.new\u002F)\n5. [cursor](https:\u002F\u002Fwww.cursor.com\u002F)\n6. [Windsurf](https:\u002F\u002Fcodeium.com\u002Fwindsurf)\n7. [cline](https:\u002F\u002Fgithub.com\u002Fcline\u002Fcline)\n8. [Trae](https:\u002F\u002Fwww.trae.ai\u002F)\n9. [MGX](https:\u002F\u002Fmgx.dev\u002F)\n10. [Roo Code](https:\u002F\u002Fgithub.com\u002FRooCodeInc\u002FRoo-Code)\n11. [Kilo Code](https:\u002F\u002Fgithub.com\u002FKilo-Org\u002Fkilocode)\n12. [AugmentCode](https:\u002F\u002Fwww.augmentcode.com\u002F)\n13. [Claude Code (`🔥`)](https:\u002F\u002Fgithub.com\u002Fanthropics\u002Fclaude-code)\n14. [Claude Code (`🔥`)](https:\u002F\u002Fgithub.com\u002Fanthropics\u002Fclaude-code) + [Happy Coder](https:\u002F\u002Fgithub.com\u002Fslopus\u002Fhappy) \u002F [CodePilot](https:\u002F\u002Fgithub.com\u002Fop7418\u002FCodePilot) \u002F [cc-connect](https:\u002F\u002Fgithub.com\u002Fchenhg5\u002Fcc-connect)\n15. [Gemini CLI](https:\u002F\u002Fgithub.com\u002Fgoogle-gemini\u002Fgemini-cli)\n16. [Serena](https:\u002F\u002Fgithub.com\u002Foraios\u002Fserena)\n17. [Claudia](https:\u002F\u002Fgithub.com\u002FgetAsterisk\u002Fclaudia)\n18. [OpenCode](https:\u002F\u002Fgithub.com\u002Fopencode-ai\u002Fopencode)\n19. [Kiro](https:\u002F\u002Fkiro.dev\u002F)\n20. [CodeBuddy](https:\u002F\u002Fcopilot.tencent.com\u002F)\n21. [Kiro](https:\u002F\u002Fkiro.dev\u002F)\n22. [CodeX (`🔥`)](https:\u002F\u002Fgithub.com\u002Fopenai\u002Fcodex)\n23. [Kimi-CLI](https:\u002F\u002Fgithub.com\u002FMoonshotAI\u002Fkimi-cli)\n24. [opencode](https:\u002F\u002Fgithub.com\u002Fanomalyco\u002Fopencode)\n25. [Multica](https:\u002F\u002Fgithub.com\u002Fmultica-ai\u002Fmultica)\n\n\n\u003Cdiv align=\"right\">\n    \u003Cb>\u003Ca href=\"#Contents\">↥ back to top\u003C\u002Fa>\u003C\u002Fb>\n\u003C\u002Fdiv>\n\n## 视频 Video\n\n#### 模型\n\n> [!NOTE]\n> 🤝[Awesome-Video-Diffusion](https:\u002F\u002Fgithub.com\u002Fshowlab\u002FAwesome-Video-Diffusion)\n\n1. [HunyuanVideo](https:\u002F\u002Fgithub.com\u002FTencent\u002FHunyuanVideo)\n2. [CogVideo](https:\u002F\u002Fgithub.com\u002FTHUDM\u002FCogVideo)\n3. [Wan2.1](https:\u002F\u002Fgithub.com\u002FWan-Video\u002FWan2.1)\n4. [Open-Sora](https:\u002F\u002Fgithub.com\u002Fhpcaitech\u002FOpen-Sora)\n5. [Open-Sora-Plan](https:\u002F\u002Fgithub.com\u002FPKU-YuanGroup\u002FOpen-Sora-Plan)\n6. [LTX-Video](https:\u002F\u002Fgithub.com\u002FLightricks\u002FLTX-Video)\n7. [Step-Video-T2V](https:\u002F\u002Fgithub.com\u002Fstepfun-ai\u002FStep-Video-T2V)\n8. [Step1X-Edit](https:\u002F\u002Fgithub.com\u002Fstepfun-ai\u002FStep1X-Edit) `Editing`\n9. [Wan2.1-VACE](https:\u002F\u002Fhuggingface.co\u002FWan-AI\u002FWan2.1-VACE-14B) `Editing`\n10. [ICEdit](https:\u002F\u002Fgithub.com\u002FRiver-Zhang\u002FICEdit) `Editing`\n11. [mochi-1-preview](https:\u002F\u002Fhuggingface.co\u002Fgenmo\u002Fmochi-1-preview)\n12. [Wan2.1-Fun](https:\u002F\u002Fhuggingface.co\u002Fcollections\u002Falibaba-pai\u002Fwan21-fun-v11-680f514c89fe7b4df9d44f17)\n13. [Wan2.1-FLF2V](https:\u002F\u002Fhuggingface.co\u002FWan-AI\u002FWan2.1-FLF2V-14B-720P) `首尾帧`\n14. [MAGI-1](https:\u002F\u002Fgithub.com\u002FSandAI-org\u002FMAGI-1) `自回归模型`\n15. [SkyReels-V2](https:\u002F\u002Fgithub.com\u002FSkyworkAI\u002FSkyReels-V2)\n16. [FramePack](https:\u002F\u002Fgithub.com\u002Flllyasviel\u002FFramePack)\n17. [Pusa-VidGen](https:\u002F\u002Fgithub.com\u002FYaofang-Liu\u002FPusa-VidGen)\n18. [Wan2.2](https:\u002F\u002Fgithub.com\u002FWan-Video\u002FWan2.2)\n19. [MoGA](https:\u002F\u002Farxiv.org\u002Fpdf\u002F2510.18692) `长视频`\n20. [LongCat-Video](https:\u002F\u002Fhuggingface.co\u002Fmeituan-longcat\u002FLongCat-Video)\n21. [HunyuanVideo-1.5](https:\u002F\u002Fgithub.com\u002FTencent-Hunyuan\u002FHunyuanVideo-1.5)\n22. [LTX-2](https:\u002F\u002Fhuggingface.co\u002FLightricks\u002FLTX-2)\n    - [Training](https:\u002F\u002Fgithub.com\u002FLightricks\u002FLTX-2\u002Fblob\u002Fmain\u002Fpackages\u002Fltx-trainer\u002FREADME.md)\n23. [daVinci-MagiHuman](https:\u002F\u002Fhuggingface.co\u002FGAIR\u002FdaVinci-MagiHuman)\n\n#### 编辑\n\n1. Wan2.1-VACE-14B: https:\u002F\u002Fhuggingface.co\u002FWan-AI\u002FWan2.1-VACE-14B\n2. Ditto: https:\u002F\u002Fgithub.com\u002FEzioBy\u002FDitto\n\n#### 训练\n\n- https:\u002F\u002Fgithub.com\u002Fhao-ai-lab\u002FFastVideo\n- https:\u002F\u002Fgithub.com\u002Ftdrussell\u002Fdiffusion-pipe\n- https:\u002F\u002Fgithub.com\u002FVideoVerses\u002FVideoTuna\n- (`🔥`) https:\u002F\u002Fgithub.com\u002Fmodelscope\u002FDiffSynth-Studio\n- https:\u002F\u002Fgithub.com\u002Fhuggingface\u002Fdiffusers\n- https:\u002F\u002Fgithub.com\u002Fkohya-ss\u002Fmusubi-tuner\n- https:\u002F\u002Fgithub.com\u002Fspacepxl\u002FHunyuanVideo-Training\n- https:\u002F\u002Fgithub.com\u002FTele-AI\u002FTeleTron\n- https:\u002F\u002Fgithub.com\u002FYaofang-Liu\u002FMochi-Full-Finetuner\n- https:\u002F\u002Fgithub.com\u002Fbghira\u002FSimpleTuner\n- https:\u002F\u002Fgithub.com\u002FX-GenGroup\u002FFlow-Factory\n\n#### 推理\n\n- https:\u002F\u002Fgithub.com\u002FModelTC\u002FLightX2V\n- https:\u002F\u002Fgithub.com\u002Fthu-ml\u002FTurboDiffusion\n\n#### 实用工具\n\n- [PySceneDetect](https:\u002F\u002Fgithub.com\u002FBreakthrough\u002FPySceneDetect): Python and OpenCV-based scene cut\u002Ftransition detection program & library.\n- [DOVER](https:\u002F\u002Fgithub.com\u002FVQAssessment\u002FDOVER): Video Quality Assessment on User Generated Contents from Aesthetic and Technical Perspectives.\n\n\u003Cdiv align=\"right\">\n    \u003Cb>\u003Ca href=\"#Contents\">↥ back to top\u003C\u002Fa>\u003C\u002Fb>\n\u003C\u002Fdiv>\n\n## 图片 Image\n\n#### 生成\n\n- [awesome-nano-banana](https:\u002F\u002Fgithub.com\u002FJimmyLv\u002Fawesome-nano-banana)\n- [Awesome-Nano-Banana-images](https:\u002F\u002Fgithub.com\u002FPicoTrex\u002FAwesome-Nano-Banana-images)\n- HunyuanImage-3.0：https:\u002F\u002Fgithub.com\u002FTencent-Hunyuan\u002FHunyuanImage-3.0\n- Seedream 4.0：https:\u002F\u002Farxiv.org\u002Fabs\u002F2509.20427\n- LongCat-Image：https:\u002F\u002Fhuggingface.co\u002Fmeituan-longcat\u002FLongCat-Image\n- Z-Image-Turbo：https:\u002F\u002Fhuggingface.co\u002FTongyi-MAI\u002FZ-Image-Turbo\n  - https:\u002F\u002Fhuggingface.co\u002FinclusionAI\u002FTwinFlow\n- Qwen-Image：https:\u002F\u002Fhuggingface.co\u002FQwen\u002FQwen-Image\n- Qwen-Image-2512：https:\u002F\u002Fhuggingface.co\u002FQwen\u002FQwen-Image-2512\n- Z-Image：https:\u002F\u002Fhuggingface.co\u002FTongyi-MAI\u002FZ-Image\n- ERNIE-Image: https:\u002F\u002Fhuggingface.co\u002Fbaidu\u002FERNIE-Image\n  - ERNIE-Image-Turbo: https:\u002F\u002Fhuggingface.co\u002FBaidu\u002FERNIE-Image-Turbo\n- Nucleus-Image: https:\u002F\u002Fhuggingface.co\u002FNucleusAI\u002FNucleus-Image\n- HiDream-O1-Image: https:\u002F\u002Fhuggingface.co\u002FHiDream-ai\u002FHiDream-O1-Image\n\n#### 编辑\n\n- ChronoEdit-14B: https:\u002F\u002Fhuggingface.co\u002Fnvidia\u002FChronoEdit-14B-Diffusers\n- Eigen-Banana-Qwen-Image-Edit: https:\u002F\u002Fhuggingface.co\u002Feigen-ai-labs\u002Feigen-banana-qwen-image-edit\n- Qwen-Image-Edit-2509: https:\u002F\u002Fhuggingface.co\u002FQwen\u002FQwen-Image-Edit-2509\n  - Upscale: https:\u002F\u002Fhuggingface.co\u002Fvafipas663\u002FQwen-Edit-2509-Upscale-LoRA\n  - Multiple-angles: https:\u002F\u002Fhuggingface.co\u002Fdx8152\u002FQwen-Edit-2509-Multiple-angles\n  - Multi-Angle-Lighting: https:\u002F\u002Fhuggingface.co\u002Fdx8152\u002FQwen-Edit-2509-Multi-Angle-Lighting\n- LongCat-Image-Edit: https:\u002F\u002Fhuggingface.co\u002Fmeituan-longcat\u002FLongCat-Image-Edit\n- Qwen-Image-Edit-2511: https:\u002F\u002Fhuggingface.co\u002FQwen\u002FQwen-Image-Edit-2511\n  - Qwen-Image-Edit-2511-Upscale2K: https:\u002F\u002Fhuggingface.co\u002Fvaliantcat\u002FQwen-Image-Edit-2511-Upscale2K\n  - Qwen-Image-Edit-2511-Multiple-Angles-LoRA: https:\u002F\u002Fhuggingface.co\u002Ffal\u002FQwen-Image-Edit-2511-Multiple-Angles-LoRA\n- FireRed-Image-Edit: https:\u002F\u002Fhuggingface.co\u002FFireRedTeam\u002FFireRed-Image-Edit-1.0\n- JoyAI-Image-Edit: https:\u002F\u002Fhuggingface.co\u002Fjdopensource\u002FJoyAI-Image-Edit\n\n#### 统一\n\n- GLM-Image: https:\u002F\u002Fhuggingface.co\u002Fzai-org\u002FGLM-Image\n- https:\u002F\u002Fhuggingface.co\u002Fblack-forest-labs\u002FFLUX.2-klein-4B\n  - https:\u002F\u002Fhuggingface.co\u002Fblack-forest-labs\u002FFLUX.2-klein-9B\n- DreamLite: https:\u002F\u002Fgithub.com\u002FByteVisionLab\u002FDreamLite\n- SenseNova-U1: https:\u002F\u002Fhuggingface.co\u002Fsensenova\u002FSenseNova-U1-8B-MoT-SFT\n\n#### 训练\n\n- Ostris：https:\u002F\u002Fgithub.com\u002Fostris\u002Fai-toolkit\n- FlymyAI：https:\u002F\u002Fgithub.com\u002FFlyMyAI\u002Fflymyai-lora-trainer\n- Nitro-T：https:\u002F\u002Fgithub.com\u002FAMD-AGI\u002FNitro-T\n- (`🔥`) DiffSynth-Studio：https:\u002F\u002Fgithub.com\u002Fmodelscope\u002FDiffSynth-Studio\n- Musubi Tuner: https:\u002F\u002Fgithub.com\u002Fkohya-ss\u002Fmusubi-tuner\n- SimpleTuner: https:\u002F\u002Fgithub.com\u002Fbghira\u002FSimpleTuner\n- MS Training: https:\u002F\u002Fwww.modelscope.cn\u002Faigc\u002FmodelTraining\n- Finetune HunyuanImage-3.0: https:\u002F\u002Fgithub.com\u002FPhotonAISG\u002Fhunyuan-image3-finetune\n- OneTrainer: https:\u002F\u002Fgithub.com\u002FNerogar\u002FOneTrainer\n- Finetune LongCat-Image and Edit: https:\u002F\u002Fgithub.com\u002Fmeituan-longcat\u002FLongCat-Image\u002Ftree\u002Fmain\u002Ftrain_examples\n- https:\u002F\u002Fgithub.com\u002FX-GenGroup\u002FFlow-Factory\n\n#### 评估\n\n- ULMEvalKit：https:\u002F\u002Fgithub.com\u002FULMEvalKit\u002FULMEvalKit\n\n#### 推理\n\n- TypemovieInfer: https:\u002F\u002Fgithub.com\u002Ftypemovie\u002FTypemovieInfer\n\n\u003Cdiv align=\"right\">\n    \u003Cb>\u003Ca href=\"#Contents\">↥ back to top\u003C\u002Fa>\u003C\u002Fb>\n\u003C\u002Fdiv>\n\n## 搜索 Search\n\n1. [OpenSearch GPT](https:\u002F\u002Fgithub.com\u002Fsupermemoryai\u002Fopensearch-ai): SearchGPT \u002F Perplexity clone, but personalised for you.\n2. [MindSearch](https:\u002F\u002Fgithub.com\u002FInternLM\u002FMindSearch): An LLM-based Multi-agent Framework of Web Search Engine (like Perplexity.ai Pro and SearchGPT).\n3. [nanoPerplexityAI](https:\u002F\u002Fgithub.com\u002FYusuke710\u002FnanoPerplexityAI): The simplest open-source implementation of perplexity.ai.\n4. [curiosity](https:\u002F\u002Fgithub.com\u002Fjank\u002Fcuriosity): Try to build a Perplexity-like user experience.\n5. [MiniPerplx](https:\u002F\u002Fgithub.com\u002Fzaidmukaddam\u002Fminiperplx): A minimalistic AI-powered search engine that helps you find information on the internet.\n\n\u003Cdiv align=\"right\">\n    \u003Cb>\u003Ca href=\"#Contents\">↥ back to top\u003C\u002Fa>\u003C\u002Fb>\n\u003C\u002Fdiv>\n\n## 语音 Speech\n\n#### TTS\n\n1. SpeechGPT-2.0-preview: https:\u002F\u002Fgithub.com\u002FOpenMOSS\u002FSpeechGPT-2.0-preview\n2. Moss-TTSD：https:\u002F\u002Fgithub.com\u002FOpenMOSS\u002FMOSS-TTSD\n3. Index-TTS：https:\u002F\u002Fgithub.com\u002Findex-tts\u002Findex-tts\n4. MegaTTS3：https:\u002F\u002Fgithub.com\u002Fbytedance\u002FMegaTTS3\n5. F5-TTS：https:\u002F\u002Fgithub.com\u002FSWivid\u002FF5-TTS\n6. GPT-SoVITS：https:\u002F\u002Fgithub.com\u002FRVC-Boss\u002FGPT-SoVITS\n7. CosyVoice：https:\u002F\u002Fgithub.com\u002FFunAudioLLM\u002FCosyVoice\n8. Spark-TTS：https:\u002F\u002Fgithub.com\u002FSparkAudio\u002FSpark-TTS\n9. OpenVoice：https:\u002F\u002Fgithub.com\u002Fmyshell-ai\u002FOpenVoice\n10. Dia：https:\u002F\u002Fgithub.com\u002Fnari-labs\u002Fdia\n11. ChatTTS：https:\u002F\u002Fgithub.com\u002F2noise\u002FChatTTS\n12. Fish Speech：https:\u002F\u002Fgithub.com\u002Ffishaudio\u002Ffish-speech\n13. Edge-TTS：https:\u002F\u002Fgithub.com\u002Frany2\u002Fedge-tts\n14. Bark：https:\u002F\u002Fgithub.com\u002Fsuno-ai\u002Fbark\n15. kokoro: https:\u002F\u002Fgithub.com\u002Fhexgrad\u002Fkokoro\n16. Higgs Audio V2: https:\u002F\u002Fgithub.com\u002Fboson-ai\u002Fhiggs-audio 【[Training](https:\u002F\u002Fgithub.com\u002FJimmyMa99\u002Ftrain-higgs-audio)】\n17. KittenTTS: https:\u002F\u002Fgithub.com\u002FKittenML\u002FKittenTTS\n18. ZipVoice: https:\u002F\u002Fgithub.com\u002Fk2-fsa\u002FZipVoice\n19. VyvoTTS: https:\u002F\u002Fgithub.com\u002FVyvo-Labs\u002FVyvoTTS\n20. VibeVoice: https:\u002F\u002Fgithub.com\u002Fmicrosoft\u002FVibeVoice\n21. Index-TTS-2: https:\u002F\u002Fhuggingface.co\u002FIndexTeam\u002FIndexTTS-2\n22. FireRedTTS2: https:\u002F\u002Fgithub.com\u002FFireRedTeam\u002FFireRedTTS2\n23. VoxCPM: https:\u002F\u002Fgithub.com\u002FOpenBMB\u002FVoxCPM\u002F\n24. Neutts-Air: https:\u002F\u002Fgithub.com\u002Fneuphonic\u002Fneutts-air\n25. Maya1: https:\u002F\u002Fhuggingface.co\u002Fmaya-research\u002Fmaya1\n26. VibeVoice: https:\u002F\u002Fhuggingface.co\u002Fcollections\u002Fmicrosoft\u002Fvibevoice\n27. GLM-TTS: https:\u002F\u002Fgithub.com\u002Fzai-org\u002FGLM-TTS\n28. Fun-CosyVoice3: https:\u002F\u002Fhuggingface.co\u002FFunAudioLLM\u002FFun-CosyVoice3-0.5B-2512\n29. Qwen3-TTS:https:\u002F\u002Fhuggingface.co\u002Fcollections\u002FQwen\u002Fqwen3-tts\n30. Ming-Omni-TTS: https:\u002F\u002Fgithub.com\u002FinclusionAI\u002FMing-omni-tts\n31. VoxCPM2: https:\u002F\u002Fhuggingface.co\u002Fopenbmb\u002FVoxCPM2\n32. OmniVoice: https:\u002F\u002Fgithub.com\u002Fk2-fsa\u002FOmniVoice\n\n#### STT\u002FASR\n\n1. Kyutai: https:\u002F\u002Fgithub.com\u002Fkyutai-labs\u002Fdelayed-streams-modeling\n2. Whisper: https:\u002F\u002Fgithub.com\u002Fopenai\u002Fwhisper\n3. Audio Flamingo 3: https:\u002F\u002Fhuggingface.co\u002Fnvidia\u002Faudio-flamingo-3\n4. Voxtral: https:\u002F\u002Fhuggingface.co\u002Fmistralai\u002FVoxtral-Mini-3B-2507\n5. Step-Audio2: https:\u002F\u002Fgithub.com\u002Fstepfun-ai\u002FStep-Audio2\n6. SoulX-Podcast: https:\u002F\u002Fhuggingface.co\u002Fcollections\u002FSoul-AILab\u002Fsoulx-podcast\n7. Omnilingual ASR: https:\u002F\u002Fgithub.com\u002Ffacebookresearch\u002Fomnilingual-asr\n8. Fun-ASR: https:\u002F\u002Fhuggingface.co\u002FFunAudioLLM\u002FFun-ASR-Nano-2512\n9. VibeVoice-ASR: https:\u002F\u002Fhuggingface.co\u002Fmicrosoft\u002FVibeVoice-ASR\n10. Qwen3-ASR: https:\u002F\u002Fgithub.com\u002FQwenLM\u002FQwen3-ASR\n\n#### Voice Interaction\n\n1. Fun-Audio-Chat: https:\u002F\u002Fhuggingface.co\u002FFunAudioLLM\u002FFun-Audio-Chat-8B\n2. Chroma 1.0: https:\u002F\u002Fhuggingface.co\u002FFlashLabs\u002FChroma-4B\n\n\u003Cdiv align=\"right\">\n    \u003Cb>\u003Ca href=\"#Contents\">↥ back to top\u003C\u002Fa>\u003C\u002Fb>\n\u003C\u002Fdiv>\n\n## 龙虾 OpenClaw\n\n1. MultiUserClaw: https:\u002F\u002Fgithub.com\u002Fjohnson7788\u002FMultiUserClaw\n2. ClawManager: https:\u002F\u002Fgithub.com\u002FYuan-lab-LLM\u002FClawManager\n3. Qclaw: https:\u002F\u002Fgithub.com\u002Fqiuzhi2046\u002FQclaw\n4. NEXU: https:\u002F\u002Fgithub.com\u002Fnexu-io\u002Fnexu\n5. OpenHanako: https:\u002F\u002Fgithub.com\u002FliliMozi\u002Fopenhanako\n\n\u003Cdiv align=\"right\">\n    \u003Cb>\u003Ca href=\"#Contents\">↥ back to top\u003C\u002Fa>\u003C\u002Fb>\n\u003C\u002Fdiv>\n\n## 统一模型 Unified Model\n\n> 现在统一模型已经从`理解+生成`变成`理解+生成+编辑`\n\n- Emu-2：https:\u002F\u002Farxiv.org\u002Fabs\u002F2312.13286\n- Emu-3：https:\u002F\u002Farxiv.org\u002Fabs\u002F2409.18869\n- Emu-1：https:\u002F\u002Farxiv.org\u002Fabs\u002F2307.05222\n- Janus：https:\u002F\u002Fgithub.com\u002Fdeepseek-ai\u002FJanus\n- Janus-Pro：http:\u002F\u002Farxiv.org\u002Fabs\u002F2508.05954\n- show-o：https:\u002F\u002Farxiv.org\u002Fabs\u002F2408.12528\n- Any-GPT：https:\u002F\u002Farxiv.org\u002Fabs\u002F2402.12226\n- Next-GPT：https:\u002F\u002Farxiv.org\u002Fpdf\u002F2309.05519.pdf\n- CoDi：https:\u002F\u002Farxiv.org\u002Fabs\u002F2305.11846\n- Seed-X：https:\u002F\u002Farxiv.org\u002Fabs\u002F2404.14396\n- Dream-LLM：https:\u002F\u002Farxiv.org\u002Fabs\u002F2309.11499\n- Chameleon：https:\u002F\u002Farxiv.org\u002Fabs\u002F2405.09818\n- Spider：https:\u002F\u002Farxiv.org\u002Fabs\u002F2411.09439\n- MedViLaM：https:\u002F\u002Farxiv.org\u002Fabs\u002F2409.19684\n- VITRON：https:\u002F\u002Fgithub.com\u002FSkyworkAI\u002FVitron\n- TokenFlow：https:\u002F\u002Fgithub.com\u002FByteFlow-AI\u002FTokenFlow\n- OneDiffusion：https:\u002F\u002Fgithub.com\u002Flehduong\u002FOneDiffusion\n- MetaMorph: https:\u002F\u002Farxiv.org\u002Fabs\u002F2412.14164\n- LlamaFusion：https:\u002F\u002Farxiv.org\u002Fabs\u002F2412.15188\n- InstructSeg：https:\u002F\u002Farxiv.org\u002Fabs\u002F2412.14006\n- VILA-U：https:\u002F\u002Farxiv.org\u002Fabs\u002F2409.04429\n- Ullava: https:\u002F\u002Fgithub.com\u002FOPPOMKLab\u002Fu-LLaVA\n- ILLUME: https:\u002F\u002Farxiv.org\u002Fabs\u002F2412.06673\n- Vitron:https:\u002F\u002Farxiv.org\u002Fabs\u002F2412.19806\n- SynerGen-VL：https:\u002F\u002Farxiv.org\u002Fabs\u002F2412.09604\n- Align Anything：https:\u002F\u002Farxiv.org\u002Fabs\u002F2412.15838\n- Mico：https:\u002F\u002Farxiv.org\u002Fabs\u002F2406.09412\n- OneLLM:https:\u002F\u002Farxiv.org\u002Fabs\u002F2312.03700\n- X-VILA:https:\u002F\u002Farxiv.org\u002Fabs\u002F2405.19335\n- OLA：https:\u002F\u002Farxiv.org\u002Fabs\u002F2502.04328\n- Transfusion: https:\u002F\u002Farxiv.org\u002Fabs\u002F2408.11039\n- JanusFlow: https:\u002F\u002Farxiv.org\u002Fabs\u002F2411.07975\n- HealthGPT：https:\u002F\u002Farxiv.org\u002Fabs\u002F2502.09838 `Medical`\n- BAGEL：https:\u002F\u002Farxiv.org\u002Fabs\u002F2505.14683\n- Qwen2.5-Omni：https:\u002F\u002Farxiv.org\u002Fabs\u002F2503.20215\n- X2I：https:\u002F\u002Farxiv.org\u002Fabs\u002F2503.06134\n- Bifrost-1：https:\u002F\u002Farxiv.org\u002Fabs\u002F2508.05954\n- OmniGen2：https:\u002F\u002Farxiv.org\u002Fabs\u002F2506.18871\n- UniPic：https:\u002F\u002Fgithub.com\u002FSkyworkAI\u002FUniPic\n- VeOmni：https:\u002F\u002Fgithub.com\u002FByteDance-Seed\u002FVeOmni `Training`\n- NextStep-1：https:\u002F\u002Farxiv.org\u002Fabs\u002F2508.10711\n- UniUGG: https:\u002F\u002Farxiv.org\u002Fabs\u002F2508.11952 `3D`\n- Omni-Video：https:\u002F\u002Farxiv.org\u002Fabs\u002F2507.06119\n- OneCAT：https:\u002F\u002Farxiv.org\u002Fabs\u002F2509.03498\n- Lumina-DiMOO：https:\u002F\u002Fgithub.com\u002FAlpha-VLLM\u002FLumina-DiMOO\n- UAE：https:\u002F\u002Fgithub.com\u002FPKU-YuanGroup\u002FUAE\n- RecA：https:\u002F\u002Farxiv.org\u002Fabs\u002F2509.07295\n- UniLM：https:\u002F\u002Farxiv.org\u002Fabs\u002F1905.03197\n- Hyper-Bagel：https:\u002F\u002Farxiv.org\u002Fabs\u002F2509.18824\n- Ming-UniVision：https:\u002F\u002Farxiv.org\u002Fabs\u002F2510.06590\n- EditVerse：https:\u002F\u002Farxiv.org\u002Fabs\u002F2509.20360\n- LightBagel: https:\u002F\u002Farxiv.org\u002Fabs\u002F2510.22946\n- DreamLLM: https:\u002F\u002Farxiv.org\u002Fabs\u002F2309.11499\n- X-Omni: https:\u002F\u002Farxiv.org\u002Fabs\u002F2507.22058\n- Ming-flash-omni-Preview: https:\u002F\u002Fhuggingface.co\u002FinclusionAI\u002FMing-flash-omni-Preview\n- Omni-View: https:\u002F\u002Farxiv.org\u002Fabs\u002F2511.07222\n- NExT-OMNI: https:\u002F\u002Farxiv.org\u002Fabs\u002F2510.13721\n- Uni-MoE-2.0-Omni: https:\u002F\u002Farxiv.org\u002Fabs\u002F2511.12609\n- LongCat-Flash-Omni: https:\u002F\u002Fhuggingface.co\u002Fmeituan-longcat\u002FLongCat-Flash-Omni\n- ShapeLLM-Omni: https:\u002F\u002Farxiv.org\u002Fabs\u002F2506.01853\n- UniGen-1.5: https:\u002F\u002Farxiv.org\u002Fabs\u002F2511.14760\n- Jodi: https:\u002F\u002Farxiv.org\u002Fabs\u002F2505.19084\n- UniModel: https:\u002F\u002Farxiv.org\u002Fabs\u002F2511.16917\n- TUNA: https:\u002F\u002Farxiv.org\u002Fabs\u002F2512.02014\n- HBridge: https:\u002F\u002Farxiv.org\u002Fabs\u002F2511.20520\n- EMMA: https:\u002F\u002Farxiv.org\u002Fabs\u002F2512.04810\n- OpenOmni: https:\u002F\u002Fgithub.com\u002FRainBowLuoCS\u002FOpenOmni\n- Ming-Flash-Omni: https:\u002F\u002Farxiv.org\u002Fabs\u002F2510.24821\n- STAR: https:\u002F\u002Fgithub.com\u002Fmm-mvr\u002Fstar\n- InternVL-U: https:\u002F\u002Fgithub.com\u002FOpenGVLab\u002FInternVL-U\n- LongCat-Next: https:\u002F\u002Fgithub.com\u002Fmeituan-longcat\u002FLongCat-Next\n- SenseNova-U1: https:\u002F\u002Fhuggingface.co\u002Fsensenova\u002FSenseNova-U1-8B-MoT-SFT\n- TUNA-2: https:\u002F\u002Fgithub.com\u002Ffacebookresearch\u002Ftuna-2\n\n\u003Cdiv align=\"right\">\n    \u003Cb>\u003Ca href=\"#Contents\">↥ back to top\u003C\u002Fa>\u003C\u002Fb>\n\u003C\u002Fdiv>\n\n## 书籍 Book\n\n1. [《大规模语言模型：从理论到实践》](https:\u002F\u002Fintro-llm.github.io\u002F)\n2. [《大语言模型》](https:\u002F\u002Fllmbook-zh.github.io\u002F)\n3. [《动手学大模型Dive into LLMs》](https:\u002F\u002Fgithub.com\u002FLordog\u002Fdive-into-llms)\n4. [《动手做AI Agent》](https:\u002F\u002Fbook.douban.com\u002Fsubject\u002F36884058\u002F)\n5. [《Build a Large Language Model (From Scratch)》](https:\u002F\u002Fgithub.com\u002Frasbt\u002FLLMs-from-scratch)\n6. [《多模态大模型》](https:\u002F\u002Fgithub.com\u002FHCPLab-SYSU\u002FBook-of-MLM)\n7. [《Generative AI Handbook: A Roadmap for Learning Resources》](https:\u002F\u002Fgenai-handbook.github.io\u002F)\n8. [《Understanding Deep Learning》](https:\u002F\u002Fudlbook.github.io\u002Fudlbook\u002F)\n9. [《Illustrated book to learn about Transformers & LLMs》](https:\u002F\u002Fwww.reddit.com\u002Fr\u002FMachineLearning\u002Fcomments\u002F1ew1hws\u002Fp_illustrated_book_to_learn_about_transformers\u002F)\n10. [《Building LLMs for Production: Enhancing LLM Abilities and Reliability with Prompting, Fine-Tuning, and RAG》](https:\u002F\u002Fwww.amazon.com\u002FBuilding-LLMs-Production-Reliability-Fine-Tuning\u002Fdp\u002FB0D4FFPFW8?crid=7OAXELUKGJE4&dib=eyJ2IjoiMSJ9.Qr3e3VSH8LSo_j1M7sV7GfS01q_W1LDYd2uGlvGJ8CW-t4DTlng6bSeOlZBryhp6HJN5K1HqWMVVgabU2wz2i9yLpy_AuaZN-raAEbenKx2NHtzZA3A4k-N7GpnldF1baCarA_V1CRF-aCdc9_3WSX7SaEzmpyDv22TTyltcKT74HAb2KiQqBGLhQS3cEAnzChcqGa1Xp-XhbMnplVwT7xZLApE3tGLhDOgi5GmSi9w.8SY_4NBEkm68YF4GwhDnz0r81ZB1d8jr-gK9IMJE5AE&dib_tag=se&keywords=building+llms+for+production&qid=1716376414&sprefix=building+llms+for+production,aps,101&sr=8-1&linkCode=sl1&tag=whatsai06-20&linkId=ee102fda07a0eb51710fcdd8b8d20c28&language=en_US&ref_=as_li_ss_tl)\n11. [《大型语言模型实战指南：应用实践与场景落地》](https:\u002F\u002Fgithub.com\u002Fliucongg\u002FLLMsBook)\n12. [《Hands-On Large Language Models》](https:\u002F\u002Fgithub.com\u002FhandsOnLLM\u002FHands-On-Large-Language-Models)\n13. [《自然语言处理：大模型理论与实践》](https:\u002F\u002Fnlp-book.swufenlp.group\u002F)\n14. [《动手学强化学习》](https:\u002F\u002Fhrl.boyuai.com\u002F)\n15. [《面向开发者的LLM入门教程》](https:\u002F\u002Fdatawhalechina.github.io\u002Fllm-cookbook\u002F#\u002F)\n16. [《大模型基础》](https:\u002F\u002Fgithub.com\u002FZJU-LLMs\u002FFoundations-of-LLMs)\n17. [Taming LLMs: A Practical Guide to LLM Pitfalls with Open Source Software ](https:\u002F\u002Fwww.tamingllms.com\u002F)\n18. [Foundations of Large Language Models](https:\u002F\u002Farxiv.org\u002Fabs\u002F2501.09223)\n19. [Textbook on reinforcement learning from human feedback](https:\u002F\u002Fgithub.com\u002Fnatolambert\u002Frlhf-book)\n20. [《大模型算法：强化学习、微调与对齐》](https:\u002F\u002Fbook.douban.com\u002Fsubject\u002F37331056\u002F)\n21. [《The Smol Training Playbook: The Secrets to Building World-Class LLMs》](https:\u002F\u002Fgithub.com\u002FWangRongsheng\u002Fawesome-LLM-resources\u002Fblob\u002Fmain\u002Fbooks\u002Fthe-smol-training-playbook-the-secrets-to-building-world-class-llms.pdf)\n22. [《从零开始构建智能体》——从零开始的智能体原理与实践教程](https:\u002F\u002Fgithub.com\u002Fdatawhalechina\u002Fhello-agents)\n23. [《Hands-On Modern RL》](https:\u002F\u002Fgithub.com\u002Fwalkinglabs\u002Fhands-on-modern-rl)\n\n\u003Cdiv align=\"right\">\n    \u003Cb>\u003Ca href=\"#Contents\">↥ back to top\u003C\u002Fa>\u003C\u002Fb>\n\u003C\u002Fdiv>\n\n## 课程 Course\n\n> [LLM Resources Hub](https:\u002F\u002Fllmresourceshub.vercel.app\u002F)\n\n1. [斯坦福 CS224N: Natural Language Processing with Deep Learning](https:\u002F\u002Fweb.stanford.edu\u002Fclass\u002Fcs224n\u002F)\n2. [吴恩达: Generative AI for Everyone](https:\u002F\u002Fwww.deeplearning.ai\u002Fcourses\u002Fgenerative-ai-for-everyone\u002F)\n3. [吴恩达: LLM series of courses](https:\u002F\u002Flearn.deeplearning.ai\u002F)\n","该项目是关于大语言模型（LLM）资源的综合汇总，涵盖了多模态生成、智能体、辅助编程、AI审稿等多个领域。其核心功能包括提供丰富的数据集、微调指南、推理方法、评估标准等，并且持续更新以保持内容的新鲜度与实用性。技术特点方面，项目涉及了多种先进的LLM技术，如Retrieval-Augmented Generation (RAG)、小语言模型及视觉语言模型等。适用于研究人员、开发者以及对人工智能特别是自然语言处理感兴趣的人士，在进行相关研究或开发时作为参考和学习资料使用。",2,"2026-06-11 03:48:36","high_star"]