[{"data":1,"prerenderedAt":1877},["ShallowReactive",2],{"header-counts":3,"footer-counts":6,"compare-fastgpt-vs-langflow":9,"compare-a-fastgpt":10,"compare-b-langflow":1239},{"tools":4,"reviews":5},78,26,{"tools":4,"reviews":5,"playbooks":7,"news":8},22,21,null,{"id":11,"title":12,"alternatives":13,"api_compatible":18,"body":20,"category":1146,"chinese_friendly":416,"cover":1147,"description":1148,"domestic":1149,"extension":1150,"faq":9,"free":1149,"github":1105,"languages":1151,"lastVerified":9,"meta":1154,"models":1155,"navigation":412,"notSuitable":1162,"opensource":412,"path":1166,"pillar":1167,"platforms":1168,"priceTable":1172,"pricing":1197,"published":1198,"relatedPlaybooks":1199,"relatedReviews":1201,"score":1206,"self_host":412,"seo":1207,"seoTitle":9,"slug":1208,"sources":1209,"stem":1220,"suitable":1221,"tagline":1227,"tags":1228,"updated":1236,"verdict":1237,"website":1098,"__hash__":1238},"tools\u002Ftools\u002Fagent\u002Fplatform\u002Ffastgpt.md","FastGPT",[14,15,16,17],"agent\u002Fplatform\u002Fdify","agent\u002Fplatform\u002Fcoze","agent\u002Fplatform\u002Flangflow","agent\u002Fplatform\u002Fn8n",[19],"openai",{"type":21,"value":22,"toc":1128},"minimark",[23,28,66,78,81,86,89,105,108,134,138,145,178,185,189,249,256,259,262,265,272,318,326,330,361,365,502,505,508,512,520,648,663,666,836,847,882,888,891,967,971,974,994,997,1015,1018,1088,1091,1116,1124],[24,25,27],"h2",{"id":26},"tldr","TL;DR",[29,30,35,55],"div",{"className":31},[32,33,34],"card","p-5","my-4",[36,37,38,42,43,50,51,54],"p",{},[39,40,41],"strong",{},"一句话："," labring 团队开源的 LLM 知识库 RAG 平台，27k+ GitHub star（截至 2026-03 数据，",[44,45,49],"a",{"href":46,"rel":47},"https:\u002F\u002Fcloud.tencent.com\u002Fdeveloper\u002Farticle\u002F2632669",[48],"nofollow","腾讯云 2026-03 教程"," 引用），Apache 2.0 许可证可商用。",[39,52,53],{},"核心优势是 RAG 链路工程做得极细","——问题预处理、混合检索、重排序、上下文组装、答案生成每一步都可视化调参。",[36,56,57,58,61,62,65],{},"最大价值在 ",[39,59,60],{},"国内企业知识库 + 私有部署"," 场景。代价是 ",[39,63,64],{},"配置门槛","：docker 基础 + 网络知识 + 一定运维能力。",[67,68,69],"blockquote",{},[36,70,71,72,77],{},"来源说明：本文基于 fastgpt.io 官方页面、github.com\u002Flabring\u002FFastGPT 仓库、",[44,73,76],{"href":74,"rel":75},"https:\u002F\u002Fwww.nanhuantech.com\u002Fzh\u002Fai-reviews\u002Ffastgpt-2025-review",[48],"南环 AI 2026-05 评测","、腾讯云开发者社区 2026-03 部署教程综合整理。版本迭代较快，命令和价格请以最新官方文档为准。",[24,79,80],{"id":80},"核心特性",[82,83,85],"h3",{"id":84},"知识库管理核心能力","知识库管理（核心能力）",[36,87,88],{},"支持文件类型：",[90,91,92,96,99,102],"ul",{},[93,94,95],"li",{},"文档：PDF \u002F Word \u002F Markdown \u002F TXT \u002F HTML",[93,97,98],{},"表格：Excel \u002F CSV",[93,100,101],{},"网页：URL 抓取 + 定时同步",[93,103,104],{},"API：通过接口推送内容",[36,106,107],{},"处理流程：上传 → 文本切分 → 向量化 → 存储 → 可用于问答。支持：",[90,109,110,116,122,128],{},[93,111,112,115],{},[39,113,114],{},"文件夹分组","：不同主题 \u002F 部门分类",[93,117,118,121],{},[39,119,120],{},"多种分块策略","：默认按段落 \u002F 按 token 数 \u002F 自定义",[93,123,124,127],{},[39,125,126],{},"批量导入","：脚本化大批量同步",[93,129,130,133],{},[39,131,132],{},"定时同步","：网页源自动更新",[82,135,137],{"id":136},"rag-流程编排最强卖点","RAG 流程编排（最强卖点）",[36,139,140,144],{},[44,141,143],{"href":74,"rel":142},[48],"南环 AI 2026 评测"," 总结的 FastGPT RAG 链路：",[146,147,148,154,160,166,172],"ol",{},[93,149,150,153],{},[39,151,152],{},"问题预处理","：改写 \u002F 扩展 \u002F 错词纠正（提升召回率）",[93,155,156,159],{},[39,157,158],{},"检索策略","：语义检索 \u002F 关键词 BM25 \u002F 混合检索，可调相似度阈值",[93,161,162,165],{},[39,163,164],{},"重排序（Rerank）","：对初步检索结果二次排序，提升相关性",[93,167,168,171],{},[39,169,170],{},"上下文组装","：最优 chunk + 问题 → prompt",[93,173,174,177],{},[39,175,176],{},"答案生成","：调大模型基于检索结果回答 + 引用标注",[36,179,180,181,184],{},"每一步都可视化调参，这是 FastGPT 比 Coze \u002F Dify 在 ",[39,182,183],{},"纯知识库 QA 精度","上更高的原因。",[82,186,188],{"id":187},"多模型支持不绑定厂商","多模型支持（不绑定厂商）",[190,191,192,205],"table",{},[193,194,195],"thead",{},[196,197,198,202],"tr",{},[199,200,201],"th",{},"模型类别",[199,203,204],{},"支持",[206,207,208,217,225,233,241],"tbody",{},[196,209,210,214],{},[211,212,213],"td",{},"国产闭源",[211,215,216],{},"豆包 \u002F 通义千问 \u002F 文心一言 \u002F 智谱 GLM \u002F Moonshot Kimi \u002F MiniMax",[196,218,219,222],{},[211,220,221],{},"开源",[211,223,224],{},"LLaMA \u002F Qwen \u002F ChatGLM \u002F DeepSeek 等可自部署",[196,226,227,230],{},[211,228,229],{},"OpenAI 系",[211,231,232],{},"GPT-5 \u002F GPT-5 mini \u002F o3",[196,234,235,238],{},[211,236,237],{},"Claude 系",[211,239,240],{},"Sonnet 4 \u002F Opus 4 \u002F Haiku",[196,242,243,246],{},[211,244,245],{},"嵌入 \u002F 重排",[211,247,248],{},"BGE \u002F m3e \u002F OpenAI text-embedding-3",[36,250,251,252,255],{},"可以在 ",[39,253,254],{},"应用级别","为不同知识库 \u002F 不同场景配置不同模型，做\"低成本 embedding + 高质量 LLM 生成\"组合。",[82,257,258],{"id":258},"工作流与高级编排",[36,260,261],{},"新版本（v4.14.x）支持类似 Dify 的工作流节点编排——条件分支、循环、HTTP 调用、代码节点。能做\"分类 → 路由到不同子知识库 → 不同模型回答\"这类复杂场景。",[82,263,264],{"id":264},"多向量库选择",[36,266,267,271],{},[44,268,270],{"href":46,"rel":269},[48],"腾讯云教程"," 公开的 4 种向量后端：",[190,273,274,284],{},[193,275,276],{},[196,277,278,281],{},[199,279,280],{},"后端",[199,282,283],{},"适用",[206,285,286,294,302,310],{},[196,287,288,291],{},[211,289,290],{},"PgVector",[211,292,293],{},"5000 万索引以下，新手 \u002F 小规模",[196,295,296,299],{},[211,297,298],{},"Milvus",[211,300,301],{},"亿级以上，高性能",[196,303,304,307],{},[211,305,306],{},"Zilliz Cloud",[211,308,309],{},"Milvus 全托管 SaaS",[196,311,312,315],{},[211,313,314],{},"SeekDB \u002F OceanBase",[211,316,317],{},"企业级国产化",[36,319,320,321,325],{},"部署时选对应 ",[322,323,324],"code",{},"docker-compose.{pgvector|milvus|...}.yml","。",[82,327,329],{"id":328},"api-与-mcp","API 与 MCP",[90,331,332,338,344,355],{},[93,333,334,337],{},[39,335,336],{},"对话 API","：流式 \u002F 非流式 HTTP，OpenAI 兼容",[93,339,340,343],{},[39,341,342],{},"知识库检索 API","：单独调检索（不走生成）做 hybrid pipeline",[93,345,346,349,350,354],{},[39,347,348],{},"MCP Server","：3005 端口暴露 MCP SSE 服务，可被 ",[44,351,353],{"href":352},"\u002Fcoding\u002Fcli\u002Fclaude-code.html","Claude Code"," 等客户端直接接入",[93,356,357,360],{},[39,358,359],{},"Webhook","：回调通知",[24,362,364],{"id":363},"部署-10-分钟docker","部署 10 分钟（Docker）",[366,367,372],"pre",{"className":368,"code":369,"language":370,"meta":371,"style":371},"language-bash shiki shiki-themes github-light github-dark","# 克隆代码\ngit clone https:\u002F\u002Fgithub.com\u002Flabring\u002FFastGPT.git\ncd FastGPT\n\n# 切到最新稳定版（参考 GitHub releases）\ngit switch -c 4.14.7.2\n\n# 选向量库版本（个人 \u002F 小规模选 pg）\ncd deploy\u002Fdocker\u002Fcn\nwget https:\u002F\u002Fdoc.fastgpt.cn\u002Fdeploy\u002Fconfig\u002Fconfig.json\n\n# 启动\ndocker-compose -f docker-compose.pg.yml up -d\n\n# 访问 http:\u002F\u002F\u003Cip>:3000，默认账号 root \u002F 1234\n","bash","",[322,373,374,383,397,407,414,420,434,439,445,453,462,467,473,491,496],{"__ignoreMap":371},[375,376,379],"span",{"class":377,"line":378},"line",1,[375,380,382],{"class":381},"sJ8bj","# 克隆代码\n",[375,384,386,390,394],{"class":377,"line":385},2,[375,387,389],{"class":388},"sScJk","git",[375,391,393],{"class":392},"sZZnC"," clone",[375,395,396],{"class":392}," https:\u002F\u002Fgithub.com\u002Flabring\u002FFastGPT.git\n",[375,398,400,404],{"class":377,"line":399},3,[375,401,403],{"class":402},"sj4cs","cd",[375,405,406],{"class":392}," FastGPT\n",[375,408,410],{"class":377,"line":409},4,[375,411,413],{"emptyLinePlaceholder":412},true,"\n",[375,415,417],{"class":377,"line":416},5,[375,418,419],{"class":381},"# 切到最新稳定版（参考 GitHub releases）\n",[375,421,423,425,428,431],{"class":377,"line":422},6,[375,424,389],{"class":388},[375,426,427],{"class":392}," switch",[375,429,430],{"class":402}," -c",[375,432,433],{"class":402}," 4.14.7.2\n",[375,435,437],{"class":377,"line":436},7,[375,438,413],{"emptyLinePlaceholder":412},[375,440,442],{"class":377,"line":441},8,[375,443,444],{"class":381},"# 选向量库版本（个人 \u002F 小规模选 pg）\n",[375,446,448,450],{"class":377,"line":447},9,[375,449,403],{"class":402},[375,451,452],{"class":392}," deploy\u002Fdocker\u002Fcn\n",[375,454,456,459],{"class":377,"line":455},10,[375,457,458],{"class":388},"wget",[375,460,461],{"class":392}," https:\u002F\u002Fdoc.fastgpt.cn\u002Fdeploy\u002Fconfig\u002Fconfig.json\n",[375,463,465],{"class":377,"line":464},11,[375,466,413],{"emptyLinePlaceholder":412},[375,468,470],{"class":377,"line":469},12,[375,471,472],{"class":381},"# 启动\n",[375,474,476,479,482,485,488],{"class":377,"line":475},13,[375,477,478],{"class":388},"docker-compose",[375,480,481],{"class":402}," -f",[375,483,484],{"class":392}," docker-compose.pg.yml",[375,486,487],{"class":392}," up",[375,489,490],{"class":402}," -d\n",[375,492,494],{"class":377,"line":493},14,[375,495,413],{"emptyLinePlaceholder":412},[375,497,499],{"class":377,"line":498},15,[375,500,501],{"class":381},"# 访问 http:\u002F\u002F\u003Cip>:3000，默认账号 root \u002F 1234\n",[36,503,504],{},"最低配置：2C4G + 20GB 硬盘 + Docker 28+ + Docker Compose 2.20+。",[36,506,507],{},"进入后台 → 账号 → 模型提供商 → 配置至少 1 个对话模型 + 1 个嵌入模型 → 即可开始建知识库。",[24,509,511],{"id":510},"云版-vs-自托管对比","云版 vs 自托管对比",[36,513,514,519],{},[44,515,518],{"href":516,"rel":517},"https:\u002F\u002Ffastgpt.io\u002Fzh\u002Fprice",[48],"fastgpt.io 官方定价"," 公开数据：",[190,521,522,550],{},[193,523,524],{},[196,525,526,529,532,535,538,541,544,547],{},[199,527,528],{},"套餐",[199,530,531],{},"价格",[199,533,534],{},"AI 积分",[199,536,537],{},"知识库索引",[199,539,540],{},"团队",[199,542,543],{},"Agent",[199,545,546],{},"知识库",[199,548,549],{},"QPM",[206,551,552,578,603,627],{},[196,553,554,557,560,563,566,569,572,575],{},[211,555,556],{},"免费",[211,558,559],{},"¥0",[211,561,562],{},"100",[211,564,565],{},"600",[211,567,568],{},"1",[211,570,571],{},"10",[211,573,574],{},"3",[211,576,577],{},"30",[196,579,580,583,586,589,592,595,598,600],{},[211,581,582],{},"基础",[211,584,585],{},"¥99\u002F月",[211,587,588],{},"4000",[211,590,591],{},"6000",[211,593,594],{},"5",[211,596,597],{},"50",[211,599,577],{},[211,601,602],{},"300",[196,604,605,608,611,614,617,619,622,624],{},[211,606,607],{},"高级",[211,609,610],{},"¥599\u002F月",[211,612,613],{},"25000",[211,615,616],{},"36000",[211,618,597],{},[211,620,621],{},"200",[211,623,562],{},[211,625,626],{},"1500",[196,628,629,632,635,638,640,642,644,646],{},[211,630,631],{},"定制",[211,633,634],{},"议价",[211,636,637],{},"弹性",[211,639,637],{},[211,641,637],{},[211,643,637],{},[211,645,637],{},[211,647,637],{},[36,649,650,653,654,657,658,662],{},[39,651,652],{},"云版适合","：不想运维、量小、要快速上线\n",[39,655,656],{},"自托管适合","：量大（10 万+ 日问答）、数据敏感、要深度定制——按 ",[44,659,661],{"href":74,"rel":660},[48],"南环评测"," 估算：\"日均 10 万次问答的企业场景，商业 SaaS 年费数十万，自建 FastGPT + 开源模型只需数万硬件投入\"",[24,664,665],{"id":665},"与同类怎么选",[190,667,668,695],{},[193,669,670],{},[196,671,672,675,677,683,689,692],{},[199,673,674],{},"维度",[199,676,12],{},[199,678,679],{},[44,680,682],{"href":681},"\u002Fagent\u002Fplatform\u002Fdify.html","Dify",[199,684,685],{},[44,686,688],{"href":687},"\u002Fagent\u002Fplatform\u002Fcoze.html","Coze",[199,690,691],{},"RAGFlow",[199,693,694],{},"AnythingLLM",[206,696,697,717,734,753,771,787,804,819],{},[196,698,699,702,705,708,711,714],{},[211,700,701],{},"核心定位",[211,703,704],{},"知识库 QA",[211,706,707],{},"综合 LLMOps",[211,709,710],{},"Bot + 工作流",[211,712,713],{},"文档解析+RAG",[211,715,716],{},"桌面级 KB",[196,718,719,721,724,726,729,731],{},[211,720,221],{},[211,722,723],{},"✅ Apache 2.0",[211,725,723],{},[211,727,728],{},"❌",[211,730,723],{},[211,732,733],{},"✅ MIT",[196,735,736,739,742,745,748,751],{},[211,737,738],{},"私有部署",[211,740,741],{},"★★★★★ docker",[211,743,744],{},"★★★★★",[211,746,747],{},"⚠️ 仅企业版",[211,749,750],{},"★★★★☆",[211,752,744],{},[196,754,755,758,761,763,766,769],{},[211,756,757],{},"RAG 深度",[211,759,760],{},"★★★★★ 最细",[211,762,750],{},[211,764,765],{},"★★★☆☆",[211,767,768],{},"★★★★★ 文档解析最强",[211,770,765],{},[196,772,773,776,778,780,782,784],{},[211,774,775],{},"工作流",[211,777,750],{},[211,779,744],{},[211,781,750],{},[211,783,765],{},[211,785,786],{},"★★☆☆☆",[196,788,789,792,795,797,800,802],{},[211,790,791],{},"上手",[211,793,794],{},"★★★☆☆ 需 docker",[211,796,750],{},[211,798,799],{},"★★★★★ 最简单",[211,801,765],{},[211,803,750],{},[196,805,806,809,811,813,815,817],{},[211,807,808],{},"中文优化",[211,810,744],{},[211,812,750],{},[211,814,744],{},[211,816,750],{},[211,818,765],{},[196,820,821,824,827,829,831,834],{},[211,822,823],{},"多平台发布",[211,825,826],{},"⚠️ API 为主",[211,828,750],{},[211,830,744],{},[211,832,833],{},"⚠️",[211,835,833],{},[36,837,838,841,842,846],{},[39,839,840],{},"怎么选","（综合 ",[44,843,845],{"href":74,"rel":844},[48],"南环 AI 评测","）：",[90,848,849,855,863,870,876],{},[93,850,851,854],{},[39,852,853],{},"核心需求是 RAG 精度"," → FastGPT",[93,856,857,860,861],{},[39,858,859],{},"需要丰富插件 + 复杂工作流 + 多平台发布"," → ",[44,862,682],{"href":681},[93,864,865,860,868],{},[39,866,867],{},"零代码、快速发布到飞书 \u002F 微信",[44,869,688],{"href":687},[93,871,872,875],{},[39,873,874],{},"文档解析（含 OCR \u002F 表格 \u002F 公式）是瓶颈"," → RAGFlow",[93,877,878,881],{},[39,879,880],{},"桌面 \u002F 单机使用"," → AnythingLLM",[36,883,884,887],{},[39,885,886],{},"很多企业同时用","：FastGPT 做知识库底座 + Coze 做前端 Bot 发布 \u002F 工作流编排。",[24,889,890],{"id":890},"避坑清单",[90,892,893,907,916,922,932,944,950,956],{},[93,894,895,898,899,902,903,906],{},[39,896,897],{},"docker-compose 镜像 tag 不一致","：",[44,900,270],{"href":46,"rel":901},[48]," 实测的坑——某些版本编排文件的 image tag 与最新 release 不一致，启动报\"镜像找不到\"，手动改 ",[322,904,905],{},"image:"," 行为正确版本即可",[93,908,909,912,913,915],{},[39,910,911],{},"3000 端口冲突","：默认占用 3000（主服务）\u002F 9000（S3 \u002F MinIO）\u002F 3005（MCP）；改 ",[322,914,478],{}," 的 ports 映射端口",[93,917,918,921],{},[39,919,920],{},"PostgreSQL pgvector 不够用就换 Milvus","：单库索引超 5000 万时 pgvector 查询性能下降，切 Milvus",[93,923,924,927,928,931],{},[39,925,926],{},"向量库选错代价大","：先评估索引量再选向量后端，迁移要重新 embedding 整库，按 ",[44,929,661],{"href":74,"rel":930},[48],"：\"新手 \u002F 小规模 PgVector，中大规模 Milvus，企业 \u002F 国产 OceanBase\"",[93,933,934,898,937,940,941],{},[39,935,936],{},"MinIO 默认密码",[322,938,939],{},"minioadmin\u002Fminioadmin","，",[39,942,943],{},"部署到公网前必须改",[93,945,946,949],{},[39,947,948],{},"分段策略影响巨大","：默认分段对长法律 \u002F 医疗文档不友好，需调\"按章节\"或\"自定义\"",[93,951,952,955],{},[39,953,954],{},"嵌入模型 ≠ 对话模型","：经常有人只配 GPT-4 没配 embedding 模型，知识库无法索引——必须同时配两类",[93,957,958,961,962,966],{},[39,959,960],{},"云版 AI 积分会过期","：未用完不能跨月累积（按 ",[44,963,965],{"href":516,"rel":964},[48],"fastgpt.io 定价 FAQ","）",[24,968,970],{"id":969},"适合-不适合","适合 \u002F 不适合",[36,972,973],{},"✅ 适合：",[90,975,976,979,982,985,988,991],{},[93,977,978],{},"企业内部知识库（员工手册 \u002F 制度 \u002F 流程）",[93,980,981],{},"产品 FAQ \u002F 用户手册问答",[93,983,984],{},"医疗 \u002F 法律 \u002F 金融垂直领域知识系统",[93,986,987],{},"数据严格不出网 + Apache 2.0 商用",[93,989,990],{},"有 docker 运维基础的技术团队",[93,992,993],{},"需要把 RAG 当后端服务的开发者（API 接入业务系统）",[36,995,996],{},"❌ 不适合：",[90,998,999,1004,1009,1012],{},[93,1000,1001,1002,966],{},"完全非技术用户（去 ",[44,1003,688],{"href":687},[93,1005,1006,1007,966],{},"主要需求是工作流 + 插件集成（去 ",[44,1008,682],{"href":681},[93,1010,1011],{},"文档解析 \u002F OCR 是首要痛点（RAGFlow）",[93,1013,1014],{},"不想自己运维 + 量很小（FastGPT 云免费版起步即可）",[24,1016,1017],{"id":1017},"相关阅读",[90,1019,1020,1033,1054,1077],{},[93,1021,1022,1023,1025,1026,1028,1029],{},"同类对比：",[44,1024,682],{"href":681}," \u002F ",[44,1027,688],{"href":687}," \u002F RAGFlow \u002F AnythingLLM \u002F ",[44,1030,1032],{"href":1031},"\u002Fagent\u002Fplatform\u002Fn8n.html","n8n",[93,1034,1035,1036,1025,1040,1025,1044,1025,1047,1025,1050],{},"概念：",[44,1037,1039],{"href":1038},"\u002Fwiki\u002Frag.html","RAG",[44,1041,1043],{"href":1042},"\u002Fwiki\u002Fembedding.html","Embedding",[44,1045,1046],{"href":1042},"Vector Database",[44,1048,1049],{"href":1038},"Reranker",[44,1051,1053],{"href":1052},"\u002Fwiki\u002Fai-agent.html","AI Agent",[93,1055,1056,1057,1025,1061,1025,1065,1025,1069,1025,1073],{},"模型：",[44,1058,1060],{"href":1059},"\u002Fmodels\u002Fdeepseek-v3.html","DeepSeek-V3",[44,1062,1064],{"href":1063},"\u002Fmodels\u002Fqwen-3.html","Qwen3",[44,1066,1068],{"href":1067},"\u002Fmodels\u002Fglm-5.2.html","GLM-5.2",[44,1070,1072],{"href":1071},"\u002Fmodels\u002Fkimi-k2.html","Kimi K2",[44,1074,1076],{"href":1075},"\u002Fmodels\u002Fdoubao-1-5-pro.html","豆包 Doubao",[93,1078,1079,1080,1025,1084],{},"进阶：",[44,1081,1083],{"href":1082},"\u002Fwiki\u002Fcontext-engineering.html","Context Engineering",[44,1085,1087],{"href":1086},"\u002Fwiki\u002Fprompt-engineering.html","Prompt Engineering",[24,1089,1090],{"id":1090},"来源",[90,1092,1093,1100,1107,1113],{},[93,1094,1095,1096],{},"官网：",[44,1097,1098],{"href":1098,"rel":1099},"https:\u002F\u002Ffastgpt.io",[48],[93,1101,1102,1103],{},"GitHub：",[44,1104,1105],{"href":1105,"rel":1106},"https:\u002F\u002Fgithub.com\u002Flabring\u002FFastGPT",[48],[93,1108,1109,1110],{},"定价：",[44,1111,516],{"href":516,"rel":1112},[48],[93,1114,1115],{},"第三方评测：南环 AI \u002F 腾讯云开发者社区 \u002F 飞书 AGI 掘金知识库",[36,1117,1118,1119,1123],{},"本卡片由 AIHO 编辑部根据官方公开资料与第三方评测整理。所有事实点均标注来源；如发现价格 \u002F 命令 \u002F 功能与最新官方信息不一致，请通过 ",[44,1120,1122],{"href":1121},"mailto:hello@aiho.net","反馈邮箱"," 反馈。",[1125,1126,1127],"style",{},"html pre.shiki code .sJ8bj, html code.shiki .sJ8bj{--shiki-default:#6A737D;--shiki-dark:#6A737D}html pre.shiki code .sScJk, html code.shiki .sScJk{--shiki-default:#6F42C1;--shiki-dark:#B392F0}html pre.shiki code .sZZnC, html code.shiki .sZZnC{--shiki-default:#032F62;--shiki-dark:#9ECBFF}html pre.shiki code .sj4cs, html code.shiki .sj4cs{--shiki-default:#005CC5;--shiki-dark:#79B8FF}html .default .shiki span {color: var(--shiki-default);background: var(--shiki-default-bg);font-style: var(--shiki-default-font-style);font-weight: var(--shiki-default-font-weight);text-decoration: var(--shiki-default-text-decoration);}html .shiki span {color: var(--shiki-default);background: var(--shiki-default-bg);font-style: var(--shiki-default-font-style);font-weight: var(--shiki-default-font-weight);text-decoration: var(--shiki-default-text-decoration);}html .dark .shiki span {color: var(--shiki-dark);background: var(--shiki-dark-bg);font-style: var(--shiki-dark-font-style);font-weight: var(--shiki-dark-font-weight);text-decoration: var(--shiki-dark-text-decoration);}html.dark .shiki span {color: var(--shiki-dark);background: var(--shiki-dark-bg);font-style: var(--shiki-dark-font-style);font-weight: var(--shiki-dark-font-weight);text-decoration: var(--shiki-dark-text-decoration);}",{"title":371,"searchDepth":399,"depth":399,"links":1129},[1130,1131,1139,1140,1141,1142,1143,1144,1145],{"id":26,"depth":385,"text":27},{"id":80,"depth":385,"text":80,"children":1132},[1133,1134,1135,1136,1137,1138],{"id":84,"depth":399,"text":85},{"id":136,"depth":399,"text":137},{"id":187,"depth":399,"text":188},{"id":258,"depth":399,"text":258},{"id":264,"depth":399,"text":264},{"id":328,"depth":399,"text":329},{"id":363,"depth":385,"text":364},{"id":510,"depth":385,"text":511},{"id":665,"depth":385,"text":665},{"id":890,"depth":385,"text":890},{"id":969,"depth":385,"text":970},{"id":1017,"depth":385,"text":1017},{"id":1090,"depth":385,"text":1090},"platform","\u002Fimg\u002Ftools\u002Ffastgpt.webp","FastGPT 真实评测：开源 LLM 知识库 RAG 平台，labring 团队出品，27k+ GitHub star。一键 docker-compose 部署、RAG 流程编排可视化、多向量库支持。AIHO 编辑部基于官方文档与社区资料整理，含与 Dify\u002FCoze 对比、避坑指南。",false,"md",[1152,1153],"zh","en",{},[1156,1157,1158,1159,1160,1161],"deepseek-v3","qwen-max","doubao-pro","gpt-4o","claude-sonnet-4","kimi",[1163,1164,1165],"完全零代码 \u002F 不懂 docker 的用户（去 Coze）","Bot 多平台一键发布场景（Coze 强项）","插件 \u002F 工作流复杂集成（去 Dify）","\u002Ftools\u002Fagent\u002Fplatform\u002Ffastgpt","agent",[1169,1170,1171],"windows","macos","linux",[1173,1178,1183,1188,1193],{"plan":1174,"price":556,"limit":1175,"cn_pay":1176,"note":1177},"Self-host 开源","全功能 + 全数据本地","—","Apache 2.0 可商用",{"plan":1179,"price":1180,"limit":1181,"cn_pay":1176,"note":1182},"云免费版","¥0\u002F月","100 AI 积分 + 600 索引 + 3 知识库","试水",{"plan":1184,"price":585,"limit":1185,"cn_pay":1186,"note":1187},"云基础版","4000 积分 + 6000 索引 + 50 Agent","✅ 微信\u002F支付宝","中小团队 SaaS",{"plan":1189,"price":610,"limit":1190,"cn_pay":1191,"note":1192},"云高级版","25000 积分 + 36000 索引 + 50 成员 + 200 Agent + 1500 QPM","✅","企业级生产",{"plan":1194,"price":634,"limit":1195,"cn_pay":1191,"note":1196},"云定制版","弹性资源 + 深度技术支持 + 专属客户经理","中大型企业","自托管开源免费 \u002F 云版 ¥0-¥599\u002F月","2026-06-18",[1200],"onboarding\u002Ffastgpt-getting-started",[1202,1203,1204,1205],"fastgpt-deep-review","coze-deep-review","coze-vs-dify","dify-deep-review",{"power":409,"ux":409,"price":416,"cn_support":416,"stability":409},{"title":12,"description":1148},"agent\u002Fplatform\u002Ffastgpt",[1210,1212,1214,1216,1218],{"title":1211,"url":1098},"FastGPT 官网",{"title":1213,"url":1105},"FastGPT GitHub",{"title":1215,"url":516},"FastGPT 定价页",{"title":1217,"url":74},"FastGPT 2025 测评（南环 AI）",{"title":1219,"url":46},"FastGPT 部署教程（腾讯云）","tools\u002Fagent\u002Fplatform\u002Ffastgpt",[1222,1223,1224,1225,1226],"企业内部知识库（员工手册、规章、流程）","产品文档智能问答（FAQ \u002F 用户手册）","垂直领域知识库（医疗、法律、金融）","数据严格不出网的合规场景","需要精细 RAG 流程编排（重排序、混合检索、阈值调节）","开源知识库问答系统，国内私有部署友好",[1229,1230,1231,1232,1233,1234,1235],"agent-platform","opensource","self-host","rag","china","knowledge-base","labring","2026-06-24","国内企业知识库私有化首选。RAG 召回工程做得很细，可视化调试好用，docker-compose 一键部署。生态插件不如 Dify 丰富。","8WByuuERL4Ll0dVQeD7A-LVHlxElsUGAZF_GhzdYohI",{"id":1240,"title":1241,"alternatives":1242,"api_compatible":9,"body":1245,"category":1146,"chinese_friendly":399,"cover":1816,"description":1817,"domestic":1149,"extension":1150,"faq":1818,"free":1149,"github":9,"languages":1831,"lastVerified":9,"meta":1833,"models":9,"navigation":412,"notSuitable":9,"opensource":412,"path":1834,"pillar":1167,"platforms":1835,"priceTable":1838,"pricing":1853,"published":1854,"relatedPlaybooks":1855,"relatedReviews":9,"score":1857,"self_host":412,"seo":1858,"seoTitle":9,"slug":16,"sources":1859,"stem":1869,"suitable":9,"tagline":1870,"tags":1871,"updated":1236,"verdict":1875,"website":1862,"__hash__":1876},"tools\u002Ftools\u002Fagent\u002Fplatform\u002Flangflow.md","Langflow",[17,1243,1244],"agent\u002Fprotocol\u002Fcomposio","agent\u002Fgeneral\u002Fopenmanus",{"type":21,"value":1246,"toc":1804},[1247,1249,1252,1255,1258,1332,1334,1360,1365,1369,1374,1400,1405,1431,1433,1490,1493,1516,1519,1661,1664,1720,1722,1748,1750,1770,1772,1802],[24,1248,27],{"id":26},[36,1250,1251],{},"Langflow 是 2023 开源、2024 被 DataStax 收购的可视化 LangChain 画布。20,000+ GitHub stars、MIT 协议、Python 实现、与 Astra DB 向量库深度集成。差异点：拖拽式画布把 LangChain primitive 映射成节点 + RAG pipeline 原生组件 + 多 agent 工作流 + 节点可下钻到 Python 代码 + 自托管 \u002F Docker \u002F DataStax Astra 云托管 + Pinecone \u002F pgvector \u002F 主流向量库适配 + Visual GUI for building LangChain pipelines。Self-host 免费 \u002F Cloud Free + Paid ~$25\u002F月起。",[36,1253,1254],{},"适合：工程团队要可视化建 LangChain 流 + 不被 SaaS 锁住；合规 \u002F 数据驻留要求自托管；要 Astra DB 一站式 RAG；想用 LangChain 但讨厌纯代码协作。不适合：业务侧 + 非工程师（用 Dify \u002F Coze）；纯无代码偏好（Langflow 仍要懂 LangChain 概念）；轻量场景 + 不需要画布（直接写 LangChain 代码）。",[24,1256,1257],{"id":1257},"核心能力",[90,1259,1260,1266,1272,1278,1284,1290,1296,1302,1308,1314,1320,1326],{},[93,1261,1262,1265],{},[39,1263,1264],{},"可视化 LangChain 画布","：节点 = primitive，连线 = 数据流",[93,1267,1268,1271],{},[39,1269,1270],{},"RAG pipeline 原生组件","：Chunking \u002F Embedding \u002F VectorStore \u002F Retriever",[93,1273,1274,1277],{},[39,1275,1276],{},"多 agent 工作流","：编排多 agent 协作",[93,1279,1280,1283],{},[39,1281,1282],{},"Python 下钻","：任意节点可写 custom Python",[93,1285,1286,1289],{},[39,1287,1288],{},"向量库适配","：Astra DB \u002F Pinecone \u002F pgvector \u002F Weaviate \u002F Chroma",[93,1291,1292,1295],{},[39,1293,1294],{},"API 部署","：流程一键导出为 REST API",[93,1297,1298,1301],{},[39,1299,1300],{},"Real-time collaboration","：多用户同 project",[93,1303,1304,1307],{},[39,1305,1306],{},"版本控制","：内置 versioning + revert",[93,1309,1310,1313],{},[39,1311,1312],{},"数据可视化","：node output \u002F data flow 可视化调试",[93,1315,1316,1319],{},[39,1317,1318],{},"角色权限","：user auth + RBAC",[93,1321,1322,1325],{},[39,1323,1324],{},"Docker \u002F pip 安装","：5 分钟启动",[93,1327,1328,1331],{},[39,1329,1330],{},"Astra-hosted cloud","：DataStax 托管选项",[24,1333,531],{"id":531},[90,1335,1336,1342,1348,1354],{},[93,1337,1338,1341],{},[39,1339,1340],{},"Self-Host (OSS)","：$0；MIT 完全免费，自付 LLM API + 服务器",[93,1343,1344,1347],{},[39,1345,1346],{},"Cloud Free","：$0；DataStax Astra-hosted 小流量",[93,1349,1350,1353],{},[39,1351,1352],{},"Cloud Paid","：~$25\u002F月起；Astra DB + 更高额度 + 团队协作",[93,1355,1356,1359],{},[39,1357,1358],{},"Enterprise","：联系销售；SSO + audit + 私有部署 + SLA",[67,1361,1362],{},[36,1363,1364],{},"自托管最低成本：$5-10\u002F月 VPS + LLM API token。Cloud Paid $25\u002F月适合不想运维的小团队。",[24,1366,1368],{"id":1367},"实测rag-应用-内部工具搭建","实测（RAG 应用 + 内部工具搭建）",[36,1370,1371],{},[39,1372,1373],{},"亮点：",[90,1375,1376,1379,1382,1385,1388,1391,1394,1397],{},[93,1377,1378],{},"画布直观，比纯写 LangChain 协作效率高 5x",[93,1380,1381],{},"节点下钻到 Python 让灵活度不被画布限制",[93,1383,1384],{},"Astra DB 集成省了配 vector store 时间",[93,1386,1387],{},"4.4\u002F5 用户评分（Propicked \u002F Tooliverse）",[93,1389,1390],{},"开源 + 自托管 + 数据驻留满足合规",[93,1392,1393],{},"多 agent 编排比裸 LangChain 调试容易",[93,1395,1396],{},"RAG pipeline 模板一键起 demo",[93,1398,1399],{},"与 DataStax 长期支持降低 abandon ware 风险",[36,1401,1402],{},[39,1403,1404],{},"踩坑：",[90,1406,1407,1410,1413,1416,1419,1422,1425,1428],{},[93,1408,1409],{},"稳定性偶发：复杂流大流量下 node 偶尔失联（SelectHub 反馈）",[93,1411,1412],{},"第三方 API 依赖：external API 失败时错误处理弱",[93,1414,1415],{},"production readiness 不算 mission-critical（要自加 observability）",[93,1417,1418],{},"LangChain 升级偶尔 break 旧 flow",[93,1420,1421],{},"文档对新组件滞后 1-2 月",[93,1423,1424],{},"中文 UI 不完整，业务侧用户上手陡",[93,1426,1427],{},"大型 flow（100+ 节点）画布卡顿",[93,1429,1430],{},"多人协作偶发同步冲突",[24,1432,791],{"id":791},[366,1434,1436],{"className":368,"code":1435,"language":370,"meta":371,"style":371},"# pip 安装\npip install langflow\nlangflow run  # http:\u002F\u002Flocalhost:7860\n\n# 或 Docker\ndocker run -p 7860:7860 langflowai\u002Flangflow:latest\n",[322,1437,1438,1443,1454,1465,1469,1474],{"__ignoreMap":371},[375,1439,1440],{"class":377,"line":378},[375,1441,1442],{"class":381},"# pip 安装\n",[375,1444,1445,1448,1451],{"class":377,"line":385},[375,1446,1447],{"class":388},"pip",[375,1449,1450],{"class":392}," install",[375,1452,1453],{"class":392}," langflow\n",[375,1455,1456,1459,1462],{"class":377,"line":399},[375,1457,1458],{"class":388},"langflow",[375,1460,1461],{"class":392}," run",[375,1463,1464],{"class":381},"  # http:\u002F\u002Flocalhost:7860\n",[375,1466,1467],{"class":377,"line":409},[375,1468,413],{"emptyLinePlaceholder":412},[375,1470,1471],{"class":377,"line":416},[375,1472,1473],{"class":381},"# 或 Docker\n",[375,1475,1476,1479,1481,1484,1487],{"class":377,"line":422},[375,1477,1478],{"class":388},"docker",[375,1480,1461],{"class":392},[375,1482,1483],{"class":402}," -p",[375,1485,1486],{"class":392}," 7860:7860",[375,1488,1489],{"class":392}," langflowai\u002Flangflow:latest\n",[36,1491,1492],{},"试 RAG 流：",[146,1494,1495,1498,1501,1504,1507,1510,1513],{},[93,1496,1497],{},"新建 flow → 选 Document QA 模板",[93,1499,1500],{},"Document Loader 节点 → 上传 PDF",[93,1502,1503],{},"Splitter → Embedder（OpenAI 或本地）",[93,1505,1506],{},"VectorStore（Astra \u002F Chroma）",[93,1508,1509],{},"Retriever + ChatOpenAI → Chat Output",[93,1511,1512],{},"部署为 API → 拿到 endpoint",[93,1514,1515],{},"复杂场景下钻节点写 Python 自定义",[24,1517,1518],{"id":1518},"对比",[190,1520,1521,1536],{},[193,1522,1523],{},[196,1524,1525,1527,1529,1531,1533],{},[199,1526,674],{},[199,1528,1241],{},[199,1530,682],{},[199,1532,1032],{},[199,1534,1535],{},"Flowise",[206,1537,1538,1555,1569,1584,1598,1614,1627,1644],{},[196,1539,1540,1543,1546,1549,1552],{},[211,1541,1542],{},"中心",[211,1544,1545],{},"LangChain primitive",[211,1547,1548],{},"LLMOps 全平台",[211,1550,1551],{},"通用 workflow",[211,1553,1554],{},"LangChain（JS）",[196,1556,1557,1559,1561,1564,1567],{},[211,1558,221],{},[211,1560,733],{},[211,1562,1563],{},"✅ AGPL",[211,1565,1566],{},"✅ Sustainable",[211,1568,733],{},[196,1570,1571,1574,1577,1580,1582],{},[211,1572,1573],{},"自托管",[211,1575,1576],{},"✅ pip\u002FDocker",[211,1578,1579],{},"✅ Docker",[211,1581,1579],{},[211,1583,1191],{},[196,1585,1586,1589,1592,1594,1596],{},[211,1587,1588],{},"可视化",[211,1590,1591],{},"✅ 旗舰",[211,1593,1191],{},[211,1595,1191],{},[211,1597,1191],{},[196,1599,1600,1603,1606,1609,1612],{},[211,1601,1602],{},"代码下钻",[211,1604,1605],{},"✅ Python",[211,1607,1608],{},"部分",[211,1610,1611],{},"✅ JS",[211,1613,1611],{},[196,1615,1616,1619,1621,1623,1625],{},[211,1617,1618],{},"RAG 内置",[211,1620,1191],{},[211,1622,1191],{},[211,1624,1608],{},[211,1626,1191],{},[196,1628,1629,1632,1635,1638,1641],{},[211,1630,1631],{},"起价（云）",[211,1633,1634],{},"$25\u002F月",[211,1636,1637],{},"$59\u002F月（Team）",[211,1639,1640],{},"自托管 $0",[211,1642,1643],{},"–",[196,1645,1646,1649,1652,1655,1658],{},[211,1647,1648],{},"适合",[211,1650,1651],{},"工程 + LangChain",[211,1653,1654],{},"业务 + LLMOps",[211,1656,1657],{},"通用自动化",[211,1659,1660],{},"JS 生态",[24,1662,1663],{"id":1663},"避坑",[90,1665,1666,1672,1678,1684,1690,1696,1702,1708,1714],{},[93,1667,1668,1671],{},[39,1669,1670],{},"自托管推荐 Docker","：pip 版本依赖冲突难调",[93,1673,1674,1677],{},[39,1675,1676],{},"生产加 observability","：langsmith \u002F sentry \u002F prometheus 必装",[93,1679,1680,1683],{},[39,1681,1682],{},"Astra DB cloud free","：起步够用，付费版起步前算清成本",[93,1685,1686,1689],{},[39,1687,1688],{},"复杂 flow 拆模块","：100+ 节点画布卡顿，拆成子 flow",[93,1691,1692,1695],{},[39,1693,1694],{},"LangChain 版本 pin","：Langflow 升级前测试 flow 兼容性",[93,1697,1698,1701],{},[39,1699,1700],{},"第三方 API 加重试","：custom Python 节点写 retry + fallback",[93,1703,1704,1707],{},[39,1705,1706],{},"多人协作 lock","：同时编辑 flow 容易冲突，加 lock 流程",[93,1709,1710,1713],{},[39,1711,1712],{},"中文场景","：UI 英文为主，业务侧用户先培训",[93,1715,1716,1719],{},[39,1717,1718],{},"Self-host vs Cloud","：合规要 self-host，省心要 Cloud",[24,1721,970],{"id":969},[90,1723,1724,1727,1730,1733,1736,1739,1742,1745],{},[93,1725,1726],{},"✅ 工程团队要可视化建 LangChain 流",[93,1728,1729],{},"✅ 合规 \u002F 数据驻留要求自托管",[93,1731,1732],{},"✅ 要 Astra DB 一站式 RAG",[93,1734,1735],{},"✅ Python 团队 + 想画布 + 想下钻代码",[93,1737,1738],{},"❌ 业务侧 + 非工程师（用 Dify \u002F Coze）",[93,1740,1741],{},"❌ 纯无代码偏好",[93,1743,1744],{},"❌ 轻量场景 + 直接写 LangChain 更快",[93,1746,1747],{},"❌ JS 生态优先（用 Flowise）",[24,1749,1017],{"id":1017},[90,1751,1752,1758,1764],{},[93,1753,1754],{},[44,1755,1757],{"href":1756},"\u002Ftools\u002Fagent\u002Fplatform\u002Fn8n","n8n 评测",[93,1759,1760],{},[44,1761,1763],{"href":1762},"\u002Ftools\u002Fagent\u002Fprotocol\u002Fcomposio","Composio 评测",[93,1765,1766],{},[44,1767,1769],{"href":1768},"\u002Ftools\u002Fagent\u002Fgeneral\u002Fopenmanus","OpenManus 评测",[24,1771,1090],{"id":1090},[146,1773,1774,1781,1788,1795],{},[93,1775,1776,1777],{},"Langflow 官网 + 定价 ",[44,1778,1779],{"href":1779,"rel":1780},"https:\u002F\u002Fwww.langflow.org\u002Fpricing",[48],[93,1782,1783,1784],{},"AutomationAtlas — Langflow $25\u002Fmo 2026 Comparison ",[44,1785,1786],{"href":1786,"rel":1787},"https:\u002F\u002Fautomationatlas.io\u002Ftools\u002Flangflow",[48],[93,1789,1790,1791],{},"xpay — LangFlow 2026 Visual AI Agent Builder（MIT + DataStax）",[44,1792,1793],{"href":1793,"rel":1794},"https:\u002F\u002Fwww.xpay.sh\u002Fresources\u002Fagentic-frameworks\u002Flangflow",[48],[93,1796,1797,1798],{},"SelectHub — LangFlow Reviews 2026（稳定性 + 生产 readiness）",[44,1799,1800],{"href":1800,"rel":1801},"https:\u002F\u002Fwww.selecthub.com\u002Fp\u002Fai-agent-builder-software\u002Flangflow\u002F",[48],[1125,1803,1127],{},{"title":371,"searchDepth":399,"depth":399,"links":1805},[1806,1807,1808,1809,1810,1811,1812,1813,1814,1815],{"id":26,"depth":385,"text":27},{"id":1257,"depth":385,"text":1257},{"id":531,"depth":385,"text":531},{"id":1367,"depth":385,"text":1368},{"id":791,"depth":385,"text":791},{"id":1518,"depth":385,"text":1518},{"id":1663,"depth":385,"text":1663},{"id":969,"depth":385,"text":970},{"id":1017,"depth":385,"text":1017},{"id":1090,"depth":385,"text":1090},"\u002Fimg\u002Ftools\u002Flangflow.webp","Langflow 真实评测：2023 开源项目，2024 被 DataStax 收购，与 Astra DB 向量数据库平台深度集成。20,000+ GitHub stars、MIT 协议。差异点：拖拽式可视化画布映射 LangChain 原语 + RAG pipeline 原生组件 + 多 agent 工作流 + Python + 自托管（pip \u002F Docker）+ DataStax Astra 托管云 + Pinecone \u002F pgvector \u002F 主流向量库适配。Cloud 免费档 + 约 $25\u002F月起。",[1819,1822,1825,1828],{"q":1820,"a":1821},"Langflow 和 Dify \u002F n8n \u002F Flowise 怎么选？","Langflow 强『可视化 LangChain 原语 + Python 代码可下钻 + 自托管 + Astra DB 集成』，工程导向。Dify 是 LLMOps 全平台（含 dataset \u002F app \u002F observability），业务侧更友好。n8n 是通用 workflow（非 LangChain 中心），70+ LangChain 节点是后加的。Flowise 是 Langflow 的同类竞品（JS 生态）。工程团队 + Python + 合规自托管 → Langflow；业务 + 完整 LLMOps → Dify；通用自动化 → n8n。",{"q":1823,"a":1824},"为什么被 DataStax 收购？","DataStax 是 Cassandra 商业公司 + Astra DB 向量数据库厂商。收购 Langflow 是为了把『可视化 LangChain builder』和『生产级 vector store』捆成一站式 RAG 解决方案。Langflow 主仓仍是 MIT 开源，但 cloud \u002F 企业版深度依赖 Astra。",{"q":1826,"a":1827},"可视化画布会不会限制灵活度？","Langflow 节点对应 LangChain primitive，可在任意节点下钻到 Python 代码 + 自定义。所以可视化层是『脚手架 + 协作工具』，不是『纯无代码黑盒』。复杂逻辑、custom tool、retriever 都能写代码扩展。",{"q":1829,"a":1830},"production readiness 如何？","用户反馈：原型 + 内部工具非常顺；大流量 \u002F 关键业务要自行加 observability \u002F 错误处理 \u002F 缓存。SelectHub 评测列出『稳定性偶发 + 第三方 API 依赖 + 非完全 production-ready』。生产部署建议 Astra-hosted cloud 或自托管 + 加 sentry \u002F langsmith \u002F prometheus。",[1153,1832],"multi",{},"\u002Ftools\u002Fagent\u002Fplatform\u002Flangflow",[1231,1836,1478,1837],"cloud","web",[1839,1843,1845,1849],{"plan":1340,"price":1840,"features":1841,"notes":1842},"$0","MIT + pip\u002FDocker + 全部组件 + Pinecone\u002Fpgvector\u002FAstra 适配","自付 LLM API + 服务器",{"plan":1346,"price":1840,"features":1844,"notes":1182},"DataStax 托管 + 小流量",{"plan":1352,"price":1846,"features":1847,"notes":1848},"~$25\u002F月起","Astra DB + 更高额度 + 更多并发 + 团队协作","按使用量阶梯",{"plan":1358,"price":1850,"features":1851,"notes":1852},"联系销售","SSO + audit + 私有部署 + SLA + 数据驻留","合规 \u002F 大客户","Self-host 完全免费 MIT \u002F Cloud Free + 付费起 ~$25·月（Astra-hosted）","2026-06-19",[1856],"onboarding\u002Frag-app-workflow",{"power":409,"ux":416,"price":416,"cn_support":399,"stability":409},{"title":1241,"description":1817},[1860,1863,1865,1867],{"name":1861,"url":1862,"accessed":1236},"Langflow 官网","https:\u002F\u002Fwww.langflow.org",{"name":1864,"url":1786,"accessed":1236},"AutomationAtlas — Langflow $25\u002Fmo 2026 Comparison",{"name":1866,"url":1793,"accessed":1236},"xpay — LangFlow 2026 Visual AI Agent Builder",{"name":1868,"url":1800,"accessed":1236},"SelectHub — LangFlow Reviews 2026","tools\u002Fagent\u002Fplatform\u002Flangflow","DataStax 出品的可视化 LangChain 画布——MIT 开源 + 20k+ stars + 自托管 + Astra DB 云托管",[1230,1872,1873,1232,1874,1458],"visual-builder","langchain","datastax","工程团队要可视化建 LangChain 流 + 不被 SaaS 锁住 + 合规要求自托管的最佳选择。要纯无代码 + 业务侧 → Dify \u002F Coze；要纯代码 → 直接写 LangChain \u002F LlamaIndex。","1rR9AipW53n2GbQ36ydZc1QwSZ0HW7kgrrg9Q0bYGKs",1784565440531]