[{"data":1,"prerenderedAt":2463},["ShallowReactive",2],{"header-counts":3,"footer-counts":6,"alt-main-ollama":8,"alt-list-ollama":583},{"tools":4,"reviews":5},98,27,{"tools":4,"reviews":5,"playbooks":7,"news":5},23,{"id":9,"title":10,"alternatives":11,"api_compatible":16,"body":31,"category":512,"chinese_friendly":499,"cover":513,"description":514,"domestic":515,"extension":516,"faq":517,"free":530,"github":531,"languages":532,"lastVerified":534,"meta":535,"models":536,"navigation":530,"notSuitable":536,"opensource":530,"path":537,"pillar":538,"platforms":539,"priceTable":544,"pricing":550,"published":551,"relatedPlaybooks":552,"relatedReviews":536,"score":555,"self_host":530,"seo":558,"seoTitle":559,"slug":560,"sources":561,"stem":569,"suitable":536,"tagline":570,"tags":571,"updated":564,"verdict":580,"website":581,"__hash__":582},"tools\u002Ftools\u002Fcoding\u002Flocal\u002Follama.md","Ollama",[12,13,14,15],"coding\u002Flocal\u002Flm-studio","coding\u002Flocal\u002Fopen-webui","coding\u002Flocal\u002Fcherry-studio","coding\u002Flocal\u002Flobe-chat",[17,18,19,20,21,22,23,24,25,26,27,28,29,10,30],"OpenAI","Anthropic","Google","Grok","Mistral","Cohere","阿里通义","百度文心","腾讯混元","Moonshot Kimi","字节豆包","DeepSeek","智谱 GLM","Hugging Face",{"type":32,"value":33,"toc":497},"minimark",[34,39,48,51,54,127,130,133,137,142,162,167,198,201,236,239,378,381,413,417,440,443,470,473],[35,36,38],"h2",{"id":37},"tldr","TL;DR",[40,41,42,43,47],"p",{},"Ollama 是本地 LLM 的 Daemon 事实标准——后台跑、暴露 REST API（11434）+ CLI、Modelfile 配置、GGUF 一站式。MIT 开源，跨 Win \u002F Mac \u002F Linux。0.19+ 起 Mac M 系列底层切 MLX 推理。模型库覆盖 Llama \u002F Qwen \u002F DeepSeek \u002F Gemma \u002F Mistral 等主流开源模型，",[44,45,46],"code",{},"ollama pull"," 一键拉。",[40,49,50],{},"适合：给 Cursor \u002F Cline \u002F Continue \u002F Open WebUI 接本地 OpenAI 兼容 endpoint、个人 \u002F 评估 \u002F 原型、嵌入应用、自动化脚本。不适合：GUI 偏好用户（用 LM Studio）、多用户并发生产服务（用 vLLM）、模型浏览 \u002F 调参界面（用 LM Studio）。",[35,52,53],{"id":53},"核心能力",[55,56,57,65,74,80,88,103,109,115,121],"ul",{},[58,59,60,64],"li",{},[61,62,63],"strong",{},"后台 Daemon","：开机自启，应用调用零延迟",[58,66,67,70,71],{},[61,68,69],{},"CLI","：",[44,72,73],{},"ollama pull \u002F run \u002F list \u002F show \u002F create \u002F serve",[58,75,76,79],{},[61,77,78],{},"Modelfile","：类 Dockerfile 注册任意 GGUF，配 SYSTEM \u002F PARAMETER \u002F TEMPLATE",[58,81,82,70,85],{},[61,83,84],{},"OpenAI 兼容 API",[44,86,87],{},"http:\u002F\u002Flocalhost:11434\u002Fv1\u002Fchat\u002Fcompletions",[58,89,90,70,93,96,97,96,100],{},[61,91,92],{},"原生 API",[44,94,95],{},"\u002Fapi\u002Fchat","、",[44,98,99],{},"\u002Fapi\u002Fgenerate",[44,101,102],{},"\u002Fapi\u002Fembeddings",[58,104,105,108],{},[61,106,107],{},"模型库","：官方注册表内置 Llama \u002F Qwen \u002F DeepSeek \u002F Gemma \u002F Mistral \u002F GPT-OSS 等",[58,110,111,114],{},[61,112,113],{},"MLX 加速（Mac）","：0.19+ 起 M 系列自动用 MLX",[58,116,117,120],{},[61,118,119],{},"量化","：默认 Q4_K_M、支持 Q5 \u002F Q8 \u002F FP16",[58,122,123,126],{},[61,124,125],{},"跨平台","：Win \u002F Mac \u002F Linux 安装包，Docker 官方镜像",[35,128,129],{"id":129},"价格",[40,131,132],{},"完全免费、MIT 开源、商用免费。",[35,134,136],{"id":135},"实测m2-pro-qwen3-coder-7b-q4","实测（M2 Pro + Qwen3-Coder-7B Q4）",[40,138,139],{},[61,140,141],{},"亮点：",[55,143,144,150,153,156,159],{},[58,145,146,149],{},[44,147,148],{},"ollama run qwen3-coder:7b"," 一行起飞，3 秒进交互",[58,151,152],{},"REST API 配 Cursor \u002F Cline \u002F Continue 几乎全工具开箱即用",[58,154,155],{},"Modelfile 写自定义编码助手（low temperature + system prompt + 16K context）几分钟搞定",[58,157,158],{},"多模型并存，按需切换，内存占用合理",[58,160,161],{},"Mac M 系列 MLX 后比旧 GGUF 模式快显著",[40,163,164],{},[61,165,166],{},"踩坑：",[55,168,169,179,185,192,195],{},[58,170,171,172,175,176],{},"默认 ",[44,173,174],{},"num_ctx"," 偏小（2048），跑长上下文要在 Modelfile 加 ",[44,177,178],{},"PARAMETER num_ctx 16384",[58,180,181,182],{},"模型默认走 0.0.0.0:11434 ↔ Docker 容器互访要 ",[44,183,184],{},"--add-host=host.docker.internal:host-gateway",[58,186,187,188,191],{},"国内 ",[44,189,190],{},"ollama.com\u002Flibrary"," 下载偶有慢，可手动 HF 下 GGUF + Modelfile 自建",[58,193,194],{},"多用户并发吞吐显著低于 vLLM",[58,196,197],{},"没有 GUI，模型浏览 \u002F 参数面板要走 LM Studio \u002F Open WebUI 配合",[35,199,200],{"id":200},"上手",[202,203,204,210,216,221,227,233],"ol",{},[58,205,206,209],{},[44,207,208],{},"curl -fsSL https:\u002F\u002Follama.ai\u002Finstall.sh | sh","（Mac \u002F Linux）；Windows winget",[58,211,212,215],{},[44,213,214],{},"ollama pull qwen3-coder:7b","（按需换模型）",[58,217,218,220],{},[44,219,148],{}," 直接聊",[58,222,223,224],{},"应用接入：baseURL = ",[44,225,226],{},"http:\u002F\u002Flocalhost:11434\u002Fv1",[58,228,229,230],{},"自定义：写 Modelfile → ",[44,231,232],{},"ollama create my-coder -f Modelfile",[58,234,235],{},"进阶：装 Open WebUI 做前端 \u002F 多人共享",[35,237,238],{"id":238},"对比",[240,241,242,263],"table",{},[243,244,245],"thead",{},[246,247,248,252,254,257,260],"tr",{},[249,250,251],"th",{},"维度",[249,253,10],{},[249,255,256],{},"LM Studio",[249,258,259],{},"vLLM",[249,261,262],{},"llama.cpp",[264,265,266,284,299,314,330,346,362],"tbody",{},[246,267,268,272,275,278,281],{},[269,270,271],"td",{},"形态",[269,273,274],{},"CLI + Daemon",[269,276,277],{},"GUI + Headless",[269,279,280],{},"Python Server",[269,282,283],{},"C++ 二进制",[246,285,286,288,291,293,296],{},[269,287,200],{},[269,289,290],{},"极低",[269,292,290],{},[269,294,295],{},"中",[269,297,298],{},"高",[246,300,301,304,306,309,312],{},[269,302,303],{},"模型浏览",[269,305,69],{},[269,307,308],{},"✅ GUI",[269,310,311],{},"无",[269,313,311],{},[246,315,316,319,322,325,328],{},[269,317,318],{},"OpenAI 兼容",[269,320,321],{},"✅ :11434",[269,323,324],{},"✅ :1234",[269,326,327],{},"✅",[269,329,327],{},[246,331,332,335,338,341,344],{},[269,333,334],{},"多用户吞吐",[269,336,337],{},"弱（~40 tok\u002Fs）",[269,339,340],{},"中（50–90）",[269,342,343],{},"强（800–12500）",[269,345,295],{},[246,347,348,351,354,356,359],{},[269,349,350],{},"MLX (Mac)",[269,352,353],{},"✅ 0.19+",[269,355,327],{},[269,357,358],{},"部分",[269,360,361],{},"–",[246,363,364,367,370,373,376],{},[269,365,366],{},"开源",[269,368,369],{},"MIT",[269,371,372],{},"闭源",[269,374,375],{},"Apache 2.0",[269,377,369],{},[35,379,380],{"id":380},"避坑",[55,382,383,389,395,401,407],{},[58,384,385,388],{},[61,386,387],{},"num_ctx 一定要设","：默认 2K 太小，跑代码 \u002F 长文档要 16K+",[58,390,391,394],{},[61,392,393],{},"Modelfile 模板别漏 TEMPLATE","：错的 chat template 会让模型输出乱码 \u002F 不停",[58,396,397,400],{},[61,398,399],{},"KV cache 爆表 = 速度悬崖","：32B 模型 32K 上下文，KV cache 可能 12+ GB，超显存自动 offload 慢 10×",[58,402,403,406],{},[61,404,405],{},"不要 0.0.0.0 直接对公网","：默认无鉴权，对外暴露走反代 + Bearer \u002F mTLS",[58,408,409,412],{},[61,410,411],{},"Mac 让它自动用 MLX","：升 0.19+；不要手动强制 GGUF + Metal",[35,414,416],{"id":415},"适合-不适合","适合 \u002F 不适合",[55,418,419,422,425,428,431,434,437],{},[58,420,421],{},"✅ 应用 \u002F IDE 接本地模型（Cursor \u002F Cline \u002F Continue）",[58,423,424],{},"✅ 个人 \u002F 评估 \u002F 脚本自动化",[58,426,427],{},"✅ Modelfile 自定义系统 prompt + 参数",[58,429,430],{},"✅ Mac M 系列 MLX 用户",[58,432,433],{},"❌ 多用户并发生产服务（用 vLLM）",[58,435,436],{},"❌ GUI 调参 \u002F 模型浏览（配 LM Studio \u002F Open WebUI）",[58,438,439],{},"❌ 极致单卡吞吐研究（直接 llama.cpp \u002F vLLM）",[35,441,442],{"id":442},"相关阅读",[55,444,445,452,458,464],{},[58,446,447],{},[448,449,451],"a",{"href":450},"\u002Ftools\u002Fcoding\u002Flocal\u002Flm-studio","LM Studio 评测",[58,453,454],{},[448,455,457],{"href":456},"\u002Ftools\u002Fcoding\u002Flocal\u002Fopen-webui","Open WebUI 评测",[58,459,460],{},[448,461,463],{"href":462},"\u002Ftools\u002Fcoding\u002Flocal\u002Fcherry-studio","Cherry Studio 评测",[58,465,466],{},[448,467,469],{"href":468},"\u002Fplaybook\u002Fonboarding\u002Frag-pipeline-build","RAG Pipeline 搭建 Playbook",[35,471,472],{"id":472},"来源",[202,474,475,483,490],{},[58,476,477,478],{},"Markaicode — Import GGUF Models into Ollama 2026（2026-05-15）",[448,479,480],{"href":480,"rel":481},"https:\u002F\u002Fmarkaicode.com\u002Fimport-gguf-models-ollama-guide",[482],"nofollow",[58,484,485,486],{},"ComputingForGeeks — Ollama Models Cheat Sheet 2026 ",[448,487,488],{"href":488,"rel":489},"https:\u002F\u002Fcomputingforgeeks.com\u002Follama-models-cheat-sheet",[482],[58,491,492,493],{},"Codersera — Ollama vs LM Studio vs vLLM vs llama.cpp vs MLX 2026 ",[448,494,495],{"href":495,"rel":496},"https:\u002F\u002Fcodersera.com\u002Fblog\u002Follama-vs-lm-studio-vs-vllm-vs-llama-cpp-vs-mlx-2026\u002F",[482],{"title":498,"searchDepth":499,"depth":499,"links":500},"",3,[501,503,504,505,506,507,508,509,510,511],{"id":37,"depth":502,"text":38},2,{"id":53,"depth":502,"text":53},{"id":129,"depth":502,"text":129},{"id":135,"depth":502,"text":136},{"id":200,"depth":502,"text":200},{"id":238,"depth":502,"text":238},{"id":380,"depth":502,"text":380},{"id":415,"depth":502,"text":416},{"id":442,"depth":502,"text":442},{"id":472,"depth":502,"text":472},"local","\u002Fimg\u002Ftools\u002Follama.webp","Ollama 真实评测：本地 LLM 的事实标准 Daemon，CLI + REST API，模型库 + Modelfile + GGUF 一站式。0.19+ 在 Mac M 系列用 MLX 加速；OpenAI 兼容端点 11434；MIT 开源 + 跨平台。",false,"md",[518,521,524,527],{"q":519,"a":520},"和 LM Studio 怎么选？","Ollama = Daemon + CLI，开机自启在 11434 端口跑，应用 \u002F IDE 调它最方便。LM Studio = GUI，模型浏览 \u002F 调参 \u002F 聊天体验更好。两者底层都基于 llama.cpp，Mac M 系列上都已切 MLX。",{"q":522,"a":523},"Modelfile 是什么？","类 Dockerfile 的模型配置：`FROM .\u002Fxxx.gguf` + PARAMETER \u002F TEMPLATE \u002F SYSTEM。把任意 GGUF 注册成本地模型供调用。`ollama create my-model -f Modelfile`。",{"q":525,"a":526},"OpenAI 兼容端点？","`http:\u002F\u002Flocalhost:11434\u002Fv1`。任何 OpenAI SDK 改 baseURL 即用。也可走原生 `\u002Fapi\u002Fchat`、`\u002Fapi\u002Fgenerate`。",{"q":528,"a":529},"并发能力？","单用户原型场景顺滑（~40 tok\u002Fs peak），多用户并发明显不如 vLLM（vLLM 的 PagedAttention + 连续批处理高 16–20×）。生产并发选 vLLM。",true,"https:\u002F\u002Fgithub.com\u002Follama\u002Follama",[533],"en","2026-08-02",{},null,"\u002Ftools\u002Fcoding\u002Flocal\u002Follama","coding",[540,541,542,543],"windows","macos","linux","docker",[545],{"plan":546,"price":547,"features":548,"notes":549},"开源版","免费","完整 CLI + REST API + Modelfile + 模型库 + MIT 协议","全平台、商用免费","完全免费 + 开源（MIT）","2026-06-19",[553,554],"onboarding\u002Frag-pipeline-build","onboarding\u002Fclaude-code-getting-started",{"power":556,"ux":556,"price":557,"cn_support":499,"stability":557},4,5,{"title":10,"description":514},"Ollama 评测 2026：本地运行大模型，开源 AI 模型管理工具，私有化部署指南","coding\u002Flocal\u002Follama",[562,565,567],{"name":563,"url":480,"accessed":564},"Markaicode — Import GGUF 2026","2026-06-24",{"name":566,"url":488,"accessed":564},"ComputingForGeeks — Ollama Cheat Sheet 2026",{"name":568,"url":495,"accessed":564},"Codersera — Ollama vs LM Studio vs vLLM 2026","tools\u002Fcoding\u002Flocal\u002Follama","本地 LLM 的 Daemon——CLI + REST API 后台跑，给 Cursor \u002F Cline \u002F Open WebUI 接本地模型最低门槛",[512,572,573,574,575,576,577,578,579],"daemon","cli","rest-api","modelfile","gguf","mlx","openai-compatible","open-source","本地 LLM 的 Daemon 事实标准，CLI \u002F Modelfile \u002F REST API 三件套配合最广泛。GUI 偏好用户走 LM Studio；多用户并发生产用 vLLM；其他场景几乎默认 Ollama。","https:\u002F\u002Follama.com","yL3ZqN3rlWsImgvBFSYlFraTqj9ki9gSJlMbXVmruyg",[584,1050,1537,1980],{"id":585,"title":256,"alternatives":586,"api_compatible":536,"body":587,"category":512,"chinese_friendly":499,"cover":1003,"description":1004,"domestic":515,"extension":516,"faq":1005,"free":530,"github":536,"languages":1018,"lastVerified":534,"meta":1020,"models":536,"navigation":530,"notSuitable":536,"opensource":515,"path":450,"pillar":538,"platforms":1021,"priceTable":1022,"pricing":1030,"published":551,"relatedPlaybooks":1031,"relatedReviews":536,"score":1032,"self_host":530,"seo":1033,"seoTitle":1034,"slug":12,"sources":1035,"stem":1041,"suitable":536,"tagline":1042,"tags":1043,"updated":564,"verdict":1047,"website":1048,"__hash__":1049},"tools\u002Ftools\u002Fcoding\u002Flocal\u002Flm-studio.md",[560,13,14,15],{"type":32,"value":588,"toc":991},[589,591,598,601,603,659,661,675,681,685,689,706,710,727,729,755,757,886,888,920,922,945,947,968,970],[35,590,38],{"id":37},[40,592,593,594,597],{},"LM Studio 是 Windows \u002F macOS \u002F Linux 桌面应用，让你像浏览 App Store 一样发现、下载、运行本地大模型（GGUF \u002F MLX 格式）。底层基于 llama.cpp + MLX，Mac M 系列原生优化。0.3+ 起新增 Headless 模式 + ",[44,595,596],{},"lms"," CLI，可在服务器跑 OpenAI 兼容 API（默认 :1234）。个人 \u002F 评估完全免费，商用咨询。",[40,599,600],{},"适合：本地 LLM 入门 \u002F 评估、Mac 用户、需要 GUI 调参 \u002F 模型比较、想给 IDE \u002F 应用接本地 OpenAI 兼容 endpoint 的开发者。不适合：多用户并发生产服务（用 vLLM）、嵌入式 \u002F 边缘部署（用 llama.cpp）、纯 CLI 工作流（用 Ollama）。",[35,602,53],{"id":53},[55,604,605,611,617,623,632,641,647,653],{},[58,606,607,610],{},[61,608,609],{},"模型浏览器","：内置 Hugging Face 检索，按 GGUF \u002F MLX \u002F 大小筛选、一键下载",[58,612,613,616],{},[61,614,615],{},"聊天界面","：System Prompt \u002F temperature \u002F top-p \u002F context size 可视化调参",[58,618,619,622],{},[61,620,621],{},"多模型并存 \u002F 切换","：同时加载多模型在不同会话中比较",[58,624,625,70,628,631],{},[61,626,627],{},"OpenAI 兼容 Local Server",[44,629,630],{},"http:\u002F\u002Flocalhost:1234\u002Fv1","，任何 SDK 即接即用",[58,633,634,70,637,640],{},[61,635,636],{},"Headless \u002F CLI",[44,638,639],{},"lms server start --port 1234","，无 GUI 可跑",[58,642,643,646],{},[61,644,645],{},"PDF \u002F 文档对话","：内置基础 RAG，丢文件就能聊",[58,648,649,652],{},[61,650,651],{},"MLX 原生支持（Mac）","：M1+ 上比 GGUF + Metal 快 30–50%",[58,654,655,658],{},[61,656,657],{},"持续批处理","：Codersera 2026 测得 50–90 tok\u002Fs（消费级 GPU + 中等模型）",[35,660,129],{"id":129},[55,662,663,669],{},[58,664,665,668],{},[61,666,667],{},"个人 \u002F 评估","：免费，全功能可用",[58,670,671,674],{},[61,672,673],{},"商用","：邮件 \u002F 官网联系 LM Studio 团队",[676,677,678],"blockquote",{},[40,679,680],{},"模型本身免费（开源权重），LM Studio 不抽水任何 token 费用。",[35,682,684],{"id":683},"实测mac-m2-pro-qwen3-coder-7b-gguf-q4_k_m","实测（Mac M2 Pro + Qwen3-Coder-7B GGUF Q4_K_M）",[40,686,687],{},[61,688,141],{},[55,690,691,694,697,700,703],{},[58,692,693],{},"模型浏览器极舒服：搜「qwen3-coder」直接列出 GGUF + MLX 各 quant，标硬件兼容度",[58,695,696],{},"加载 7B Q4 模型 \u003C 3 秒，生成 ~75 tok\u002Fs",[58,698,699],{},"Local Server 开了 Cursor 直接接 baseURL → 本地代码补全零成本",[58,701,702],{},"MLX 版同模型 ~110 tok\u002Fs，差距显著",[58,704,705],{},"多窗口加载 2 个模型并排测，调 prompt 直观",[40,707,708],{},[61,709,166],{},[55,711,712,715,718,721,724],{},[58,713,714],{},"模型库依赖 Hugging Face，国内访问要镜像 \u002F 代理",[58,716,717],{},"GPU 显存吃满后会自动 offload 到 CPU，无提示就慢下来",[58,719,720],{},"Headless 模式相对 Ollama 偏新，文档稍少",[58,722,723],{},"闭源应用（虽免费），不适合企业合规挂钩",[58,725,726],{},"中文 UI 可用但部分菜单仍英文",[35,728,200],{"id":200},[202,730,731,734,737,740,743,750],{},[58,732,733],{},"lmstudio.ai 下载（Mac \u002F Windows \u002F Linux）",[58,735,736],{},"打开 → Discover 标签 → 搜模型（如 qwen3-coder、deepseek-v3 GGUF\u002FMLX）→ Download",[58,738,739],{},"Chat 标签 → 选模型 → 调参聊天",[58,741,742],{},"Local Server 标签 → Start Server → 默认端口 1234",[58,744,745,746,749],{},"在你的应用里：",[44,747,748],{},"baseURL = \"http:\u002F\u002Flocalhost:1234\u002Fv1\"","，API Key 任意",[58,751,752,753],{},"Headless：",[44,754,639],{},[35,756,238],{"id":238},[240,758,759,774],{},[243,760,761],{},[246,762,763,765,767,769,772],{},[249,764,251],{},[249,766,256],{},[249,768,10],{},[249,770,771],{},"Open WebUI",[249,773,262],{},[264,775,776,792,807,821,833,845,859,872],{},[246,777,778,780,783,786,789],{},[269,779,271],{},[269,781,782],{},"GUI + CLI",[269,784,785],{},"CLI Daemon",[269,787,788],{},"Docker UI",[269,790,791],{},"二进制",[246,793,794,796,799,802,804],{},[269,795,303],{},[269,797,798],{},"✅ 内置",[269,800,801],{},"CLI pull",[269,803,311],{},[269,805,806],{},"手动",[246,808,809,812,814,817,819],{},[269,810,811],{},"参数调优 GUI",[269,813,327],{},[269,815,816],{},"❌",[269,818,358],{},[269,820,816],{},[246,822,823,825,827,829,831],{},[269,824,84],{},[269,826,324],{},[269,828,321],{},[269,830,327],{},[269,832,327],{},[246,834,835,837,839,841,843],{},[269,836,350],{},[269,838,327],{},[269,840,353],{},[269,842,361],{},[269,844,361],{},[246,846,847,850,853,855,857],{},[269,848,849],{},"多用户并发",[269,851,852],{},"弱",[269,854,852],{},[269,856,327],{},[269,858,295],{},[246,860,861,863,866,868,870],{},[269,862,366],{},[269,864,865],{},"闭源（免费）",[269,867,369],{},[269,869,369],{},[269,871,369],{},[246,873,874,877,879,882,884],{},[269,875,876],{},"上手难度",[269,878,290],{},[269,880,881],{},"低",[269,883,295],{},[269,885,298],{},[35,887,380],{"id":380},[55,889,890,896,902,908,914],{},[58,891,892,895],{},[61,893,894],{},"国内下模型走镜像","：HF 直连慢 \u002F 卡，配 HF_ENDPOINT=hf-mirror.com",[58,897,898,901],{},[61,899,900],{},"显存爆 ≠ 报错","：GPU 装不下会无声 offload 到 CPU，关注生成速度，必要时降 quant 或换小模型",[58,903,904,907],{},[61,905,906],{},"MLX 优先（Mac M 系列）","：能下 MLX 版就别下 GGUF，速度差距明显",[58,909,910,913],{},[61,911,912],{},"Local Server 暴露要谨慎","：默认 0.0.0.0 + 无鉴权，对外开放前加反代 + Bearer",[58,915,916,919],{},[61,917,918],{},"闭源合规要核","：企业内部使用前查 license；商用必须联系官方",[35,921,416],{"id":415},[55,923,924,927,930,933,936,939,942],{},[58,925,926],{},"✅ 本地 LLM 入门 \u002F 评估",[58,928,929],{},"✅ Mac M 系列用户",[58,931,932],{},"✅ 想给 Cursor \u002F Cline 接本地 OpenAI 兼容 endpoint",[58,934,935],{},"✅ 需要 GUI 调参 \u002F 模型比较",[58,937,938],{},"❌ 多用户并发生产服务",[58,940,941],{},"❌ 嵌入式 \u002F 边缘设备",[58,943,944],{},"❌ 强合规 \u002F 必须开源审计",[35,946,442],{"id":442},[55,948,949,954,958,962],{},[58,950,951],{},[448,952,953],{"href":537},"Ollama 评测",[58,955,956],{},[448,957,457],{"href":456},[58,959,960],{},[448,961,463],{"href":462},[58,963,964],{},[448,965,967],{"href":966},"\u002Fplaybook\u002Fonboarding\u002Fclaude-code-getting-started","Claude Code 上手 Playbook",[35,969,472],{"id":472},[202,971,972,979,986],{},[58,973,974,975],{},"LM Studio 官网 ",[448,976,977],{"href":977,"rel":978},"https:\u002F\u002Flmstudio.ai\u002F",[482],[58,980,981,982],{},"Codersera — LM Studio Complete Guide 2026 ",[448,983,984],{"href":984,"rel":985},"https:\u002F\u002Fcodersera.com\u002Fblog\u002Flm-studio-complete-guide-2026\u002F",[482],[58,987,492,988],{},[448,989,495],{"href":495,"rel":990},[482],{"title":498,"searchDepth":499,"depth":499,"links":992},[993,994,995,996,997,998,999,1000,1001,1002],{"id":37,"depth":502,"text":38},{"id":53,"depth":502,"text":53},{"id":129,"depth":502,"text":129},{"id":683,"depth":502,"text":684},{"id":200,"depth":502,"text":200},{"id":238,"depth":502,"text":238},{"id":380,"depth":502,"text":380},{"id":415,"depth":502,"text":416},{"id":442,"depth":502,"text":442},{"id":472,"depth":502,"text":472},"\u002Fimg\u002Ftools\u002Flm-studio.webp","LM Studio 真实评测：跨平台桌面应用，运行本地 GGUF \u002F MLX 大模型。50–90 tok\u002Fs 持续批处理、OpenAI 兼容本地 API（默认端口 1234）、Headless 模式、Mac \u002F Win 双端。对个人开发者免费，企业咨询。",[1006,1009,1012,1015],{"q":1007,"a":1008},"和 Ollama 怎么选？","LM Studio 是 GUI 优先（模型浏览器 + 参数面板 + 聊天界面），适合个人 \u002F 评估 \u002F 上手。Ollama 是 CLI \u002F Daemon 优先（后台跑 + REST API），适合应用嵌入 \u002F 脚本调用。两者都基于 llama.cpp，在 Mac M 系列上都已用 MLX。",{"q":1010,"a":1011},"支持 MLX 吗？","支持。Mac M1+ 上可加载 MLX 格式模型，速度比 GGUF + Metal 快 30–50%。模型搜索时筛选 MLX 即可。",{"q":1013,"a":1014},"OpenAI 兼容 API 怎么用？","开 Local Server → 默认端口 1234 → `http:\u002F\u002Flocalhost:1234\u002Fv1`。任何 OpenAI SDK 把 baseURL 改这个就能跑本地模型，零代码改动。",{"q":1016,"a":1017},"Headless 模式？","0.3+ 起支持 `lms server start` CLI 启动后台服务，无 GUI 即可跑 OpenAI 兼容 API，适合服务器 \u002F SSH 场景。",[533,1019],"zh",{},[540,541,542],[1023,1026],{"plan":667,"price":547,"features":1024,"notes":1025},"全功能 GUI + Headless API + GGUF\u002FMLX","供个人 \u002F 评估使用",{"plan":673,"price":1027,"features":1028,"notes":1029},"联系咨询","团队部署 \u002F 商用 license","邮件 \u002F 官网联系","免费（个人 \u002F 评估） \u002F 企业 \u002F 商用咨询",[553,554],{"power":556,"ux":557,"price":557,"cn_support":499,"stability":556},{"title":256,"description":1004},"LM Studio 评测 2026：本地运行开源大模型，图形化界面，AI 模型管理",[1036,1038,1040],{"name":1037,"url":977,"accessed":564},"LM Studio 官网",{"name":1039,"url":984,"accessed":564},"Codersera — LM Studio Complete Guide 2026",{"name":568,"url":495,"accessed":564},"tools\u002Fcoding\u002Flocal\u002Flm-studio","本地 LLM 的 GUI 首选——模型浏览器 + GGUF\u002FMLX 推理 + OpenAI 兼容 API + Mac 原生优化",[512,1044,576,577,1045,1046,578],"gui","llama-cpp","mac","Mac \u002F Windows 桌面本地 LLM 的 GUI 首选——上手最快、模型浏览最舒服、自带 OpenAI 兼容 API。批量服务 \u002F 多用户场景用 vLLM；纯 CLI \u002F 嵌入应用走 Ollama。","https:\u002F\u002Flmstudio.ai","Pj3jNb1Z4S55e91UkW1ZMgI86ps_Wm7yCc7zeXPF05U",{"id":1051,"title":771,"alternatives":1052,"api_compatible":1053,"body":1054,"category":512,"chinese_friendly":556,"cover":1486,"description":1487,"domestic":515,"extension":516,"faq":1488,"free":530,"github":1501,"languages":1502,"lastVerified":534,"meta":1503,"models":536,"navigation":530,"notSuitable":536,"opensource":530,"path":456,"pillar":538,"platforms":1504,"priceTable":1506,"pricing":1515,"published":551,"relatedPlaybooks":1516,"relatedReviews":536,"score":1517,"self_host":530,"seo":1518,"seoTitle":1519,"slug":13,"sources":1520,"stem":1527,"suitable":536,"tagline":1528,"tags":1529,"updated":564,"verdict":1534,"website":1535,"__hash__":1536},"tools\u002Ftools\u002Fcoding\u002Flocal\u002Fopen-webui.md",[15,14,560,12],[17,18,19,20,21,22,23,24,25,26,27,28,29,10,30],{"type":32,"value":1055,"toc":1474},[1056,1058,1061,1064,1066,1128,1130,1133,1137,1141,1169,1173,1197,1199,1233,1235,1357,1359,1402,1404,1427,1429,1449,1451],[35,1057,38],{"id":37},[40,1059,1060],{},"Open WebUI（原 Ollama WebUI）是 MIT 开源、自托管 AI 平台，最常见用法是 Docker 跑起来给 Ollama 套一个 ChatGPT 风格前端。GitHub 126k+ stars、282M+ Docker pulls，事实上的本地 AI 前端首选。支持任意 OpenAI 兼容后端 + RAG 知识库 + 多用户账号 + 工具调用 + MCP-OpenAPI 代理 + 联网搜索 + 语音 + 图像生成。",[40,1062,1063],{},"适合：团队 \u002F 家庭 \u002F 公司部署一份共享、要 Web 端访问、多用户分账号、SearXNG 联网搜索、Confluence \u002F S3 \u002F GitHub 数据源同步。不适合：单人桌面体验（用 Cherry Studio）、零运维 \u002F 不愿碰 Docker。",[35,1065,53],{"id":53},[55,1067,1068,1074,1080,1086,1092,1098,1104,1110,1116,1122],{},[58,1069,1070,1073],{},[61,1071,1072],{},"多模型后端","：Ollama \u002F OpenAI \u002F vLLM \u002F Anthropic \u002F Groq \u002F LocalAI \u002F 任意 OpenAI 兼容",[58,1075,1076,1079],{},[61,1077,1078],{},"多用户 + RBAC","：注册 \u002F 邀请 \u002F 角色权限 \u002F 工作区隔离",[58,1081,1082,1085],{},[61,1083,1084],{},"RAG 知识库","：上传文档 \u002F 网址 \u002F SearXNG 联网搜索 → 向量化 → 对话引用",[58,1087,1088,1091],{},[61,1089,1090],{},"Tools \u002F Functions","：Python 写函数即扩展（联网 \u002F 计算器 \u002F 自定义 API）",[58,1093,1094,1097],{},[61,1095,1096],{},"mcpo","：MCP-to-OpenAPI 代理，任意 MCP 服务器接进来",[58,1099,1100,1103],{},[61,1101,1102],{},"oikb","：知识库同步本地文件夹 \u002F GitHub \u002F S3 \u002F Confluence 等 40+ 源",[58,1105,1106,1109],{},[61,1107,1108],{},"open-terminal \u002F cptr","：给 AI 真实终端 + 文件 + 沙箱执行",[58,1111,1112,1115],{},[61,1113,1114],{},"图像生成","：Stable Diffusion \u002F DALL·E \u002F 自托管接入",[58,1117,1118,1121],{},[61,1119,1120],{},"语音输入 \u002F TTS","：内置",[58,1123,1124,1127],{},[61,1125,1126],{},"企业 LTS","：custom branding + SLA + 长期支持版本（联系销售）",[35,1129,129],{"id":129},[40,1131,1132],{},"完全免费、MIT 开源、商用免费。Enterprise 提供品牌定制 + SLA + LTS。",[35,1134,1136],{"id":1135},"实测ubuntu-2404-ollama-后端-5-人小团队","实测（Ubuntu 24.04 + Ollama 后端 + 5 人小团队）",[40,1138,1139],{},[61,1140,141],{},[55,1142,1143,1150,1153,1160,1163,1166],{},[58,1144,1145,1146,1149],{},"单条 ",[44,1147,1148],{},"docker run"," 五分钟上线",[58,1151,1152],{},"自带的多用户 + 角色权限省去重新搭 Auth",[58,1154,1155,1156,1159],{},"RAG 直传 30 个 PDF 后向量化顺利，对话中 ",[44,1157,1158],{},"#知识库"," 引用准确",[58,1161,1162],{},"mcpo 把 GitHub MCP 服务器接进来，团队对话里直接 issue \u002F PR 操作",[58,1164,1165],{},"模型切换流畅，OpenAI + Ollama 并存",[58,1167,1168],{},"SearXNG 联网搜索给模型实时信息，过时知识截止问题缓解",[40,1170,1171],{},[61,1172,166],{},[55,1174,1175,1178,1184,1191,1194],{},[58,1176,1177],{},"Docker 镜像 ~1.5GB，首次拉取偏慢",[58,1179,171,1180,1183],{},[44,1181,1182],{},"0.0.0.0"," 公网暴露要加 HTTPS + 反代",[58,1185,1186,1187,1190],{},"嵌入模型 ",[44,1188,1189],{},"sentence-transformers"," 中文效果一般，建议换 bge-m3",[58,1192,1193],{},"多用户共享 Ollama 时并发吞吐瓶颈在 Ollama，不在 Open WebUI（生产用 vLLM 后端）",[58,1195,1196],{},"版本升级要看 changelog，部分 minor 含 breaking 改动",[35,1198,200],{"id":200},[202,1200,1201,1207,1214,1217,1220,1223,1226],{},[58,1202,1203,1204],{},"装 Docker → ",[44,1205,1206],{},"docker run -d -p 3000:8080 --add-host=host.docker.internal:host-gateway -v open-webui:\u002Fapp\u002Fbackend\u002Fdata --name open-webui --restart always ghcr.io\u002Fopen-webui\u002Fopen-webui:main",[58,1208,1209,1210,1213],{},"浏览器开 ",[44,1211,1212],{},"http:\u002F\u002Flocalhost:3000"," → 注册第一个账号（管理员）",[58,1215,1216],{},"设置 → Connections → 连接 Ollama \u002F 加 OpenAI Key",[58,1218,1219],{},"Models → Pull \u002F Discover 模型",[58,1221,1222],{},"Workspaces → 建知识库 → 上传文档",[58,1224,1225],{},"Tools → 启用 \u002F 写自定义函数",[58,1227,1228,1229,1232],{},"生产部署：Nginx 反代 + Let's Encrypt + 备份 ",[44,1230,1231],{},"\u002Fapp\u002Fbackend\u002Fdata"," volume",[35,1234,238],{"id":238},[240,1236,1237,1253],{},[243,1238,1239],{},[246,1240,1241,1243,1245,1248,1251],{},[249,1242,251],{},[249,1244,771],{},[249,1246,1247],{},"LobeChat",[249,1249,1250],{},"Cherry Studio",[249,1252,256],{},[264,1254,1255,1270,1284,1298,1312,1327,1343],{},[246,1256,1257,1259,1262,1265,1268],{},[269,1258,271],{},[269,1260,1261],{},"Docker \u002F 桌面",[269,1263,1264],{},"Web + 桌面",[269,1266,1267],{},"桌面",[269,1269,1267],{},[246,1271,1272,1275,1278,1280,1282],{},[269,1273,1274],{},"多用户",[269,1276,1277],{},"✅ 一等",[269,1279,327],{},[269,1281,311],{},[269,1283,311],{},[246,1285,1286,1289,1292,1294,1296],{},[269,1287,1288],{},"RAG",[269,1290,1291],{},"✅ 强 + oikb",[269,1293,327],{},[269,1295,327],{},[269,1297,852],{},[246,1299,1300,1303,1306,1308,1310],{},[269,1301,1302],{},"工具 \u002F MCP",[269,1304,1305],{},"✅ mcpo",[269,1307,327],{},[269,1309,327],{},[269,1311,852],{},[246,1313,1314,1317,1320,1323,1325],{},[269,1315,1316],{},"自托管",[269,1318,1319],{},"✅ Docker \u002F K8s",[269,1321,1322],{},"✅ Docker",[269,1324,311],{},[269,1326,311],{},[246,1328,1329,1332,1335,1338,1341],{},[269,1330,1331],{},"GitHub Stars",[269,1333,1334],{},"126k+",[269,1336,1337],{},"72k+",[269,1339,1340],{},"60k+",[269,1342,361],{},[246,1344,1345,1348,1350,1352,1355],{},[269,1346,1347],{},"开源协议",[269,1349,369],{},[269,1351,369],{},[269,1353,1354],{},"AGPL-3.0",[269,1356,865],{},[35,1358,380],{"id":380},[55,1360,1361,1367,1375,1384,1390,1396],{},[58,1362,1363,1366],{},[61,1364,1365],{},"不要裸 0.0.0.0 + HTTP 暴露公网","：默认无 HTTPS，必上反代 + 强密码 + 速率限制",[58,1368,1369,1374],{},[61,1370,1371,1372,1232],{},"备份 ",[44,1373,1231],{},"：知识库 \u002F 用户 \u002F 对话全在里面",[58,1376,1377,1380,1381,1383],{},[61,1378,1379],{},"中文 RAG 换嵌入模型","：默认 ",[44,1382,1189],{}," 中文一般，配 bge-m3 或硅基流动嵌入 API",[58,1385,1386,1389],{},[61,1387,1388],{},"mcpo 工具范围谨慎","：MCP 给 AI 真实能力，第三方服务器审一遍",[58,1391,1392,1395],{},[61,1393,1394],{},"后端吞吐看 Ollama","：5+ 并发上 vLLM 后端，Ollama 单 worker 会排队",[58,1397,1398,1401],{},[61,1399,1400],{},"升级前看 changelog","：weekly 更新，偶有 breaking",[35,1403,416],{"id":415},[55,1405,1406,1409,1412,1415,1418,1421,1424],{},[58,1407,1408],{},"✅ 团队 \u002F 家庭 \u002F 公司多人共享 AI 平台",[58,1410,1411],{},"✅ 要 Web 端访问 \u002F 移动端兼容",[58,1413,1414],{},"✅ 自托管 \u002F 完全控制数据",[58,1416,1417],{},"✅ MCP \u002F 工具调用刚需",[58,1419,1420],{},"❌ 单人桌面体验（用 Cherry Studio）",[58,1422,1423],{},"❌ 零运维 \u002F 不愿碰 Docker",[58,1425,1426],{},"❌ iOS 原生 App 主力",[35,1428,442],{"id":442},[55,1430,1431,1437,1441,1445],{},[58,1432,1433],{},[448,1434,1436],{"href":1435},"\u002Ftools\u002Fcoding\u002Flocal\u002Flobe-chat","LobeChat 评测",[58,1438,1439],{},[448,1440,463],{"href":462},[58,1442,1443],{},[448,1444,953],{"href":537},[58,1446,1447],{},[448,1448,469],{"href":468},[35,1450,472],{"id":472},[202,1452,1453,1460,1467],{},[58,1454,1455,1456],{},"Open WebUI 官方文档 ",[448,1457,1458],{"href":1458,"rel":1459},"https:\u002F\u002Fdocs.openwebui.com\u002F",[482],[58,1461,1462,1463],{},"Local AI Master — Open WebUI Setup Guide 2026 ",[448,1464,1465],{"href":1465,"rel":1466},"https:\u002F\u002Flocalaimaster.com\u002Fblog\u002Fopen-webui-setup-guide",[482],[58,1468,1469,1470],{},"AIToolDiscovery — Set Up Open-WebUI with Ollama 2026 ",[448,1471,1472],{"href":1472,"rel":1473},"https:\u002F\u002Fwww.aitooldiscovery.com\u002Fhow-to\u002Fsetup-open-webui-ollama",[482],{"title":498,"searchDepth":499,"depth":499,"links":1475},[1476,1477,1478,1479,1480,1481,1482,1483,1484,1485],{"id":37,"depth":502,"text":38},{"id":53,"depth":502,"text":53},{"id":129,"depth":502,"text":129},{"id":1135,"depth":502,"text":1136},{"id":200,"depth":502,"text":200},{"id":238,"depth":502,"text":238},{"id":380,"depth":502,"text":380},{"id":415,"depth":502,"text":416},{"id":442,"depth":502,"text":442},{"id":472,"depth":502,"text":472},"\u002Fimg\u002Ftools\u002Fopen-webui.webp","Open WebUI 2026 真实评测：MIT 开源、自托管 ChatGPT 替代和 Ollama Web 前端。支持 Docker 一行部署、Ollama\u002FOpenAI\u002FvLLM 多后端、RAG 知识库、多用户、联网搜索、工具调用和 MCP-to-OpenAPI，适合团队私有 AI 平台。",[1489,1492,1495,1498],{"q":1490,"a":1491},"Docker 一行命令真的够用吗？","够。`docker run -d -p 3000:8080 --add-host=host.docker.internal:host-gateway -v open-webui:\u002Fapp\u002Fbackend\u002Fdata --name open-webui --restart always ghcr.io\u002Fopen-webui\u002Fopen-webui:main`，5 分钟可上线、能多人注册、能接 Ollama \u002F OpenAI。生产再加反代 + HTTPS + 备份。",{"q":1493,"a":1494},"支持哪些模型后端？","Ollama（首选）+ 任何 OpenAI 兼容 endpoint：OpenAI 官方 \u002F Anthropic（OpenAI 兼容代理）\u002F vLLM \u002F Groq \u002F LocalAI \u002F 自建 baseURL。可同时配多个，对话中切换。",{"q":1496,"a":1497},"RAG \u002F 知识库怎么做？","内置：上传 PDF \u002F DOCX \u002F TXT、网址抓取、SearXNG 联网搜索 → 自动向量化 → 在对话中 `#` 引用知识库。配套 oikb 项目可同步本地文件夹 \u002F GitHub \u002F S3 \u002F Confluence 等 40+ 数据源。",{"q":1499,"a":1500},"MCP 怎么接？","通过 mcpo（官方的 MCP-to-OpenAPI 代理）把任意 MCP 服务器暴露成 OpenAPI 工具，再在 Open WebUI 注册即可。无需写 glue code。","https:\u002F\u002Fgithub.com\u002Fopen-webui\u002Fopen-webui",[533,1019],{},[543,542,541,540,1505],"kubernetes",[1507,1510],{"plan":546,"price":547,"features":1508,"notes":1509},"全功能 \u002F 多用户 \u002F RAG \u002F Tools \u002F 联网搜索 \u002F MCP-OpenAPI 代理 \u002F Docker \u002F K8s","MIT 协议",{"plan":1511,"price":1512,"features":1513,"notes":1514},"Enterprise","咨询","Custom branding \u002F SLA \u002F LTS 长期支持版本","邮件官方","完全免费（MIT 开源） \u002F Enterprise SLA 联系",[553,554],{"power":557,"ux":556,"price":557,"cn_support":556,"stability":557},{"title":771,"description":1487},"Open WebUI 评测 2026：自托管 ChatGPT 替代，Ollama 前端部署指南",[1521,1523,1525],{"name":1522,"url":1458,"accessed":564},"Open WebUI 官方文档",{"name":1524,"url":1465,"accessed":564},"Local AI Master — Open WebUI Setup Guide 2026",{"name":1526,"url":1472,"accessed":564},"AIToolDiscovery — Open-WebUI with Ollama 2026","tools\u002Fcoding\u002Flocal\u002Fopen-webui","自托管的 ChatGPT 替代：Ollama \u002F OpenAI 兼容、多用户、RAG、126k+ GitHub stars",[512,1530,543,1531,1532,1533,579],"self-host","rag","multi-user","ollama","自托管多用户 AI 前端的事实标准。团队 \u002F 家庭 \u002F 公司部署一份共享，多模型聚合 + RAG + 工具调用全有。单机 \u002F 桌面体验首选 Cherry Studio \u002F LobeChat。","https:\u002F\u002Fdocs.openwebui.com","5Y-zy_XMSVV_gsS0aa6V7zk268r2TvaIcssIp-HXM_U",{"id":1538,"title":1250,"alternatives":1539,"api_compatible":536,"body":1540,"category":512,"chinese_friendly":557,"cover":1929,"description":1930,"domestic":530,"extension":516,"faq":1931,"free":530,"github":1944,"languages":1945,"lastVerified":534,"meta":1946,"models":536,"navigation":530,"notSuitable":536,"opensource":530,"path":462,"pillar":538,"platforms":1947,"priceTable":1949,"pricing":1957,"published":551,"relatedPlaybooks":1958,"relatedReviews":536,"score":1960,"self_host":530,"seo":1961,"seoTitle":1962,"slug":14,"sources":1963,"stem":1970,"suitable":536,"tagline":1971,"tags":1972,"updated":564,"verdict":1977,"website":1978,"__hash__":1979},"tools\u002Ftools\u002Fcoding\u002Flocal\u002Fcherry-studio.md",[15,12,560,13],{"type":32,"value":1541,"toc":1917},[1542,1544,1547,1550,1552,1596,1598,1610,1615,1619,1623,1640,1644,1661,1663,1687,1689,1813,1815,1847,1849,1872,1874,1892,1894],[35,1543,38],{"id":37},[40,1545,1546],{},"Cherry Studio 是一款开源、跨平台（Windows \u002F macOS \u002F Linux \u002F Android）的桌面 AI 客户端，定位『全能 AI 工作台』：把 OpenAI \u002F Anthropic \u002F Google \u002F DeepSeek 等云端模型，以及 Ollama \u002F LM Studio 本地模型，全部聚合到同一个桌面应用里管理。内置 300+ 助手模板、本地 RAG 知识库、Markdown + Mermaid 渲染、MCP 协议支持，所有对话数据本地存储 + WebDAV 备份。AGPL-3.0 开源、GitHub 60k+ stars，企业版可联系商务做私有化部署。",[40,1548,1549],{},"适合：中文 AI 重度用户、想统一管理多家模型、需要本地知识库 RAG、关注数据本地存储的开发者 \u002F 研究者。不适合：要 Web 端访问 \u002F Docker 自托管 \u002F 团队多人共享 \u002F iOS 端使用。",[35,1551,53],{"id":53},[55,1553,1554,1560,1566,1572,1578,1584,1590],{},[58,1555,1556,1559],{},[61,1557,1558],{},"多模型聚合","：OpenAI \u002F Claude \u002F Gemini \u002F DeepSeek \u002F Qwen \u002F Kimi \u002F Moonshot 等云端 + Ollama \u002F LM Studio 本地",[58,1561,1562,1565],{},[61,1563,1564],{},"本地 RAG 知识库","：拖拽 PDF \u002F Word \u002F Excel \u002F PPT \u002F 网址 \u002F sitemap → 自动向量化 → 检索增强问答 + 来源追溯",[58,1567,1568,1571],{},[61,1569,1570],{},"300+ 助手模板","：编程 \u002F 写作 \u002F 翻译 \u002F 学习 \u002F 角色扮演开箱即用，可自定义 System Prompt",[58,1573,1574,1577],{},[61,1575,1576],{},"MCP 协议","：扩展工具调用 \u002F 联网搜索 \u002F 文件操作",[58,1579,1580,1583],{},[61,1581,1582],{},"数据本地优先","：对话历史本地存储，WebDAV 同步，不上传第三方",[58,1585,1586,1589],{},[61,1587,1588],{},"多模态","：图片识别 \u002F PDF 阅读 \u002F Markdown + Mermaid + 代码高亮",[58,1591,1592,1595],{},[61,1593,1594],{},"AI 绘画 + 翻译","：内置主流 SD \u002F DALL·E \u002F 翻译 API 集成",[35,1597,129],{"id":129},[55,1599,1600,1605],{},[58,1601,1602,1604],{},[61,1603,546],{},"：完全免费，AGPL-3.0",[58,1606,1607,1609],{},[61,1608,1511],{},"：私有化部署 + 团队协作 + 资源管控，联系销售",[676,1611,1612],{},[40,1613,1614],{},"模型 API 费用按你自己绑定的供应商计费；本地 Ollama \u002F LM Studio 零成本。",[35,1616,1618],{"id":1617},"实测mac-m2-中型知识库","实测（Mac M2 + 中型知识库）",[40,1620,1621],{},[61,1622,141],{},[55,1624,1625,1628,1631,1634,1637],{},[58,1626,1627],{},"中文 UI \u002F 文档 \u002F 社区都顶级，零门槛上手",[58,1629,1630],{},"本地 RAG 拖入 30+ PDF 后向量化 \u003C 2 分钟（用 bge-m3）",[58,1632,1633],{},"多模型并排回答：让 Claude \u002F GPT \u002F DeepSeek 同回一个问题做比较",[58,1635,1636],{},"MCP 接 Brave Search + 自定义工具流畅",[58,1638,1639],{},"WebDAV 同步坚果云 \u002F 阿里云盘，桌面 + 移动设备数据互通",[40,1641,1642],{},[61,1643,166],{},[55,1645,1646,1649,1652,1655,1658],{},[58,1647,1648],{},"没有 Web 端 \u002F Docker 自托管（要这个用 LobeChat）",[58,1650,1651],{},"iOS 版尚未发布（roadmap 中）",[58,1653,1654],{},"大型 PDF（>100 MB）向量化偶有失败，要切小",[58,1656,1657],{},"助手市场质量参差，要自筛",[58,1659,1660],{},"模型 API 调用全靠你自己付费，新手要先理解 API Key 概念",[35,1662,200],{"id":200},[202,1664,1665,1668,1671,1678,1681,1684],{},[58,1666,1667],{},"cherry-ai.com 下载客户端（或 GitHub releases）",[58,1669,1670],{},"设置 → 模型服务 → 填 OpenAI \u002F Claude \u002F DeepSeek API Key",[58,1672,1673,1674],{},"（可选）本地：装 Ollama → Cherry Studio 自动识别 endpoint ",[448,1675,1676],{"href":1676,"rel":1677},"http:\u002F\u002Flocalhost:11434",[482],[58,1679,1680],{},"新建知识库 → 拖文件 \u002F 加网址 → 等向量化",[58,1682,1683],{},"新对话 → 选模型 → 勾知识库 → 提问",[58,1685,1686],{},"进阶：自定义助手（System Prompt）+ MCP 扩展工具",[35,1688,238],{"id":238},[240,1690,1691,1705],{},[243,1692,1693],{},[246,1694,1695,1697,1699,1701,1703],{},[249,1696,251],{},[249,1698,1250],{},[249,1700,1247],{},[249,1702,256],{},[249,1704,771],{},[264,1706,1707,1719,1733,1747,1760,1774,1789,1801],{},[246,1708,1709,1711,1713,1715,1717],{},[269,1710,271],{},[269,1712,1267],{},[269,1714,1264],{},[269,1716,1267],{},[269,1718,1261],{},[246,1720,1721,1723,1726,1728,1731],{},[269,1722,1558],{},[269,1724,1725],{},"✅ 云 + 本地",[269,1727,1725],{},[269,1729,1730],{},"本地为主",[269,1732,1725],{},[246,1734,1735,1738,1741,1743,1745],{},[269,1736,1737],{},"知识库 RAG",[269,1739,1740],{},"✅ 强",[269,1742,1740],{},[269,1744,852],{},[269,1746,327],{},[246,1748,1749,1752,1754,1756,1758],{},[269,1750,1751],{},"MCP",[269,1753,327],{},[269,1755,327],{},[269,1757,852],{},[269,1759,327],{},[246,1761,1762,1765,1768,1770,1772],{},[269,1763,1764],{},"自托管 \u002F Web",[269,1766,1767],{},"无 Web",[269,1769,1322],{},[269,1771,311],{},[269,1773,1322],{},[246,1775,1776,1779,1782,1784,1787],{},[269,1777,1778],{},"中文",[269,1780,1781],{},"5\u002F5",[269,1783,1781],{},[269,1785,1786],{},"4\u002F5",[269,1788,1786],{},[246,1790,1791,1793,1795,1797,1799],{},[269,1792,1347],{},[269,1794,1354],{},[269,1796,369],{},[269,1798,865],{},[269,1800,369],{},[246,1802,1803,1805,1807,1809,1811],{},[269,1804,1331],{},[269,1806,1340],{},[269,1808,1337],{},[269,1810,361],{},[269,1812,1334],{},[35,1814,380],{"id":380},[55,1816,1817,1823,1829,1835,1841],{},[58,1818,1819,1822],{},[61,1820,1821],{},"API Key 别明文外泄","：客户端配置文件以明文存 Key，机器借出前先清；团队共享用企业版 \u002F 自建中转",[58,1824,1825,1828],{},[61,1826,1827],{},"知识库别一次塞太多","：单库 1000+ 文档检索质量明显下降，按主题切分多个知识库",[58,1830,1831,1834],{},[61,1832,1833],{},"嵌入模型选择","：免费 bge-m3 够用；专业用付费 Pro\u002FBAAI\u002Fbge-m3 或 OpenAI text-embedding-3",[58,1836,1837,1840],{},[61,1838,1839],{},"WebDAV 同步先小范围测","：知识库向量数据较大，先备份对话再开同步",[58,1842,1843,1846],{},[61,1844,1845],{},"MCP 工具来源要可控","：MCP 是给 AI 真实工具能力，第三方插件审一遍代码",[35,1848,416],{"id":415},[55,1850,1851,1854,1857,1860,1863,1866,1869],{},[58,1852,1853],{},"✅ 中文用户、AI 重度使用 \u002F 多模型管理",[58,1855,1856],{},"✅ 需要本地 RAG 知识库",[58,1858,1859],{},"✅ 关注数据隐私 \u002F 本地存储",[58,1861,1862],{},"✅ 想用 Ollama \u002F LM Studio 本地模型",[58,1864,1865],{},"❌ 需要 Web 端 \u002F Docker 自托管",[58,1867,1868],{},"❌ 团队多人共享 \u002F SSO",[58,1870,1871],{},"❌ iOS 主力用户",[35,1873,442],{"id":442},[55,1875,1876,1880,1884,1888],{},[58,1877,1878],{},[448,1879,1436],{"href":1435},[58,1881,1882],{},[448,1883,451],{"href":450},[58,1885,1886],{},[448,1887,953],{"href":537},[58,1889,1890],{},[448,1891,469],{"href":468},[35,1893,472],{"id":472},[202,1895,1896,1903,1910],{},[58,1897,1898,1899],{},"Cherry Studio 官网（功能 + 下载）",[448,1900,1901],{"href":1901,"rel":1902},"https:\u002F\u002Fwww.cherry-ai.com\u002F",[482],[58,1904,1905,1906],{},"MBLUO Studio — Cherry Studio 评测 2026 ",[448,1907,1908],{"href":1908,"rel":1909},"https:\u002F\u002Fmbluostudio.com\u002Ftools\u002Fcherry-studio",[482],[58,1911,1912,1913],{},"Cursor IDE 博客 — Cherry Studio 完全指南（2025-03）",[448,1914,1915],{"href":1915,"rel":1916},"https:\u002F\u002Fwww.cursor-ide.com\u002Fblog\u002Fcherry-studio-guide",[482],{"title":498,"searchDepth":499,"depth":499,"links":1918},[1919,1920,1921,1922,1923,1924,1925,1926,1927,1928],{"id":37,"depth":502,"text":38},{"id":53,"depth":502,"text":53},{"id":129,"depth":502,"text":129},{"id":1617,"depth":502,"text":1618},{"id":200,"depth":502,"text":200},{"id":238,"depth":502,"text":238},{"id":380,"depth":502,"text":380},{"id":415,"depth":502,"text":416},{"id":442,"depth":502,"text":442},{"id":472,"depth":502,"text":472},"\u002Fimg\u002Ftools\u002Fcherry-studio.webp","Cherry Studio 真实评测：开源跨平台桌面 AI 客户端，集成 OpenAI \u002F Claude \u002F Gemini \u002F DeepSeek + Ollama \u002F LM Studio 本地模型，内置 300+ 助手模板 + 本地 RAG 知识库。AGPL-3.0 开源、GitHub 60k+ stars，企业版另询。",[1932,1935,1938,1941],{"q":1933,"a":1934},"Cherry Studio 真的免费吗？","是。客户端完全免费、AGPL-3.0 开源，模型调用走你自己的 API Key（OpenAI \u002F Claude \u002F Gemini \u002F DeepSeek 等付费）或本地 Ollama \u002F LM Studio（零成本）。",{"q":1936,"a":1937},"本地知识库怎么用？","在『知识库』面板新建，拖文件 \u002F 加网址 \u002F 填 sitemap，系统自动向量化（默认 BAAI\u002Fbge-m3 或硅基流动的 Pro 版）；提问时勾选要检索的知识库，AI 会基于检索片段答题并标出来源。",{"q":1939,"a":1940},"和 LobeChat 怎么选？","都开源、多模型、有 RAG。LobeChat 是 Web + 桌面双形态，可自托管 Docker，72k stars；Cherry Studio 是纯桌面（Win\u002FMac\u002FLinux\u002FAndroid），不支持 Web 部署但桌面体验更精细，60k+ stars。要 Web 访问 \u002F 公司多人共享选 LobeChat；个人重度选 Cherry Studio。",{"q":1942,"a":1943},"支持 MCP \u002F 插件吗？","支持 MCP（Model Context Protocol）扩展，配合自定义助手（System Prompt）可扩展工具调用、联网搜索等能力。","https:\u002F\u002Fgithub.com\u002FCherryHQ\u002Fcherry-studio",[1019,533],{},[540,541,542,1948],"android",[1950,1953],{"plan":546,"price":547,"features":1951,"notes":1952},"300+ 助手模板 \u002F 云端 + 本地模型 \u002F 知识库 \u002F MCP \u002F WebDAV 备份","AGPL-3.0 开源",{"plan":1511,"price":1954,"features":1955,"notes":1956},"联系销售","私有化部署 \u002F 团队协作 \u002F AI 资源管控 \u002F 知识库管理","面向企业团队","开源免费 \u002F 企业版联系销售",[553,1959],"onboarding\u002Fcursor-mcp-deep-integration",{"power":556,"ux":557,"price":557,"cn_support":557,"stability":556},{"title":1250,"description":1930},"Cherry Studio 评测 2026：AI 客户端工具，多模型桌面助手，开源免费",[1964,1966,1968],{"name":1965,"url":1901,"accessed":564},"Cherry Studio 官网",{"name":1967,"url":1908,"accessed":564},"MBLUO Studio — Cherry Studio 评测",{"name":1969,"url":1915,"accessed":564},"Cursor IDE 博客 — Cherry Studio 指南","tools\u002Fcoding\u002Flocal\u002Fcherry-studio","全能 AI 客户端：多模型聚合 + 本地知识库 + 300+ 助手模板，跨平台桌面应用",[512,1973,1974,1975,1531,579,1976],"desktop","multi-model","knowledge-base","china","国产 AI 桌面客户端第一梯队，多模型聚合 + 本地 RAG + 中文体验顶级。需要 Web 部署 \u002F 自托管选 LobeChat；只要桌面体验完整选 Cherry Studio。","https:\u002F\u002Fcherry-ai.com","nhG3iaQ6G7dAWmMarVHUYX3EU47pSjb4yoUqlqZy18k",{"id":1981,"title":1247,"alternatives":1982,"api_compatible":1983,"body":1984,"category":512,"chinese_friendly":557,"cover":2417,"description":2418,"domestic":530,"extension":516,"faq":2419,"free":530,"github":2389,"languages":2432,"lastVerified":534,"meta":2433,"models":536,"navigation":530,"notSuitable":536,"opensource":530,"path":1435,"pillar":538,"platforms":2434,"priceTable":2436,"pricing":2443,"published":551,"relatedPlaybooks":2444,"relatedReviews":536,"score":2445,"self_host":530,"seo":2446,"seoTitle":2447,"slug":15,"sources":2448,"stem":2455,"suitable":536,"tagline":2456,"tags":2457,"updated":564,"verdict":2460,"website":2461,"__hash__":2462},"tools\u002Ftools\u002Fcoding\u002Flocal\u002Flobe-chat.md",[14,13,560,12],[17,18,19,20,21,22,23,24,25,26,27,28,29,10,30],{"type":32,"value":1985,"toc":2405},[1986,1988,1995,1998,2000,2062,2064,2077,2080,2084,2088,2108,2112,2129,2131,2157,2159,2292,2294,2332,2334,2360,2362,2380,2382],[35,1987,38],{"id":37},[40,1989,1990,1991,1994],{},"LobeChat 是 LobeHub 团队的开源 AI 聊天框架，2023 年发布、GitHub 72k+ stars、MIT 协议。",[61,1992,1993],{},"Web + 桌面 + Docker 自托管三形态","，把 OpenAI \u002F Claude \u002F Gemini \u002F DeepSeek \u002F Qwen \u002F Kimi \u002F Ollama \u002F LM Studio 等 80+ 模型聚合到一个现代设计的客户端里。内置 RAG 知识库 + 插件市场 + 助手市场 + 多模型对比 + MCP，是当下综合最强的多模型 AI 客户端之一。",[40,1996,1997],{},"适合：需要 Web 端访问、Docker 自托管、多模型对比、丰富助手市场的用户；中文重度用户；想给团队 \u002F 家庭部署一个共享 AI 工作台。不适合：只用桌面 + 不需要 Web（Cherry Studio 同样优秀且更精细）、强企业 RBAC + 多租户（Open WebUI 多用户更完善）。",[35,1999,53],{"id":53},[55,2001,2002,2007,2013,2018,2024,2030,2035,2040,2046,2052],{},[58,2003,2004,2006],{},[61,2005,1558],{},"：OpenAI \u002F Claude \u002F Gemini \u002F DeepSeek \u002F Qwen \u002F Kimi \u002F 豆包 \u002F Groq \u002F Together \u002F OpenRouter \u002F Ollama \u002F LM Studio",[58,2008,2009,2012],{},[61,2010,2011],{},"多模型对比","：同 prompt 给多模型并排回答",[58,2014,2015,2017],{},[61,2016,1564],{},"：上传 PDF \u002F Word \u002F 网页 → 向量化 → 检索引用",[58,2019,2020,2023],{},[61,2021,2022],{},"插件市场","：联网搜索 \u002F 代码执行 \u002F 图像生成 \u002F 翻译等几十款官方插件",[58,2025,2026,2029],{},[61,2027,2028],{},"助手市场","：几百个预设 AI 角色，一键导入",[58,2031,2032,2034],{},[61,2033,1576],{},"：扩展任意工具能力",[58,2036,2037],{},[61,2038,2039],{},"代码解释器 \u002F 文件上传 \u002F TTS \u002F 多模态",[58,2041,2042,2045],{},[61,2043,2044],{},"Web + 桌面 + Docker","：三形态，数据可完全本地",[58,2047,2048,2051],{},[61,2049,2050],{},"LobeHub Cloud","：官方云托管，免部署",[58,2053,2054,2057,2058,2061],{},[61,2055,2056],{},"快捷指令 \u002F 工作流","：自定义 prompt 模板，",[44,2059,2060],{},"\u002Fpodcast-summary"," 类用法",[35,2063,129],{"id":129},[55,2065,2066,2072],{},[58,2067,2068,2071],{},[61,2069,2070],{},"自托管 \u002F 桌面","：完全免费、MIT 开源",[58,2073,2074,2076],{},[61,2075,2050],{},"：订阅制，云端托管 + 团队协作 + 同步",[40,2078,2079],{},"模型 API 费用按你自己的供应商付费；本地 Ollama \u002F LM Studio 零成本。",[35,2081,2083],{"id":2082},"实测m2-自托管-docker连-openai-deepseek-本地-ollama","实测（M2 + 自托管 Docker，连 OpenAI + DeepSeek + 本地 Ollama）",[40,2085,2086],{},[61,2087,141],{},[55,2089,2090,2093,2096,2099,2102,2105],{},[58,2091,2092],{},"界面颜值是这一类工具里第一档（深色 \u002F 透明 \u002F 现代感）",[58,2094,2095],{},"多模型并排对比对选型极其有用：写一道复杂题，Claude \u002F GPT \u002F DeepSeek 直接对比答案",[58,2097,2098],{},"知识库 RAG 上传 50+ PDF 后检索准确，引用片段可视化",[58,2100,2101],{},"助手市场拿来即用——「Code Reviewer」「Translation Polish」节省 prompt 编写",[58,2103,2104],{},"Docker 一键部署，团队 5 人共享流畅",[58,2106,2107],{},"多平台数据同步（Cloud \u002F WebDAV）",[40,2109,2110],{},[61,2111,166],{},[55,2113,2114,2117,2120,2123,2126],{},[58,2115,2116],{},"自托管要熟悉 Docker + 反代 + HTTPS",[58,2118,2119],{},"国内连 OpenAI \u002F Claude 需自带网络方案",[58,2121,2122],{},"Web 版数据存 LobeHub，隐私敏感场景走桌面 \u002F Docker",[58,2124,2125],{},"插件市场质量参差，要自筛",[58,2127,2128],{},"团队多人共享需配 LobeHub Cloud 或自建数据库（Postgres + S3）",[35,2130,200],{"id":200},[202,2132,2133,2136,2142,2145,2148,2151,2154],{},[58,2134,2135],{},"选形态：Web（chat.lobehub.com 注册即用） \u002F 桌面（GitHub Releases 下载） \u002F Docker",[58,2137,2138,2139],{},"Docker：",[44,2140,2141],{},"docker run -d -p 3210:3210 -e OPENAI_API_KEY=sk-xxx --name lobe-chat lobehub\u002Flobe-chat",[58,2143,2144],{},"设置 → AI 服务商 → 添加 OpenAI \u002F Claude \u002F DeepSeek \u002F Ollama",[58,2146,2147],{},"模型选择器测试对话",[58,2149,2150],{},"知识库：拖文件 → 等向量化 → 对话引用",[58,2152,2153],{},"助手市场拉「Code Reviewer」「论文翻译润色」试用",[58,2155,2156],{},"进阶：插件市场启用联网搜索 \u002F 代码执行；MCP 自定义工具",[35,2158,238],{"id":238},[240,2160,2161,2175],{},[243,2162,2163],{},[246,2164,2165,2167,2169,2171,2173],{},[249,2166,251],{},[249,2168,1247],{},[249,2170,1250],{},[249,2172,771],{},[249,2174,256],{},[264,2176,2177,2189,2202,2214,2227,2243,2255,2268,2280],{},[246,2178,2179,2181,2183,2185,2187],{},[269,2180,271],{},[269,2182,2044],{},[269,2184,1267],{},[269,2186,1261],{},[269,2188,1267],{},[246,2190,2191,2193,2196,2198,2200],{},[269,2192,1558],{},[269,2194,2195],{},"✅ 80+",[269,2197,327],{},[269,2199,327],{},[269,2201,1730],{},[246,2203,2204,2206,2208,2210,2212],{},[269,2205,2011],{},[269,2207,1277],{},[269,2209,327],{},[269,2211,852],{},[269,2213,852],{},[246,2215,2216,2218,2220,2222,2225],{},[269,2217,1737],{},[269,2219,327],{},[269,2221,327],{},[269,2223,2224],{},"✅ + oikb",[269,2226,852],{},[246,2228,2229,2232,2235,2238,2241],{},[269,2230,2231],{},"插件 \u002F 助手市场",[269,2233,2234],{},"✅ 丰富",[269,2236,2237],{},"300+ 助手",[269,2239,2240],{},"Tools",[269,2242,852],{},[246,2244,2245,2247,2249,2251,2253],{},[269,2246,1751],{},[269,2248,327],{},[269,2250,327],{},[269,2252,1305],{},[269,2254,852],{},[246,2256,2257,2259,2262,2264,2266],{},[269,2258,1274],{},[269,2260,2261],{},"配 Cloud \u002F 自建",[269,2263,311],{},[269,2265,1277],{},[269,2267,311],{},[246,2269,2270,2272,2274,2276,2278],{},[269,2271,1331],{},[269,2273,1337],{},[269,2275,1340],{},[269,2277,1334],{},[269,2279,361],{},[246,2281,2282,2284,2286,2288,2290],{},[269,2283,1347],{},[269,2285,369],{},[269,2287,1354],{},[269,2289,369],{},[269,2291,865],{},[35,2293,380],{"id":380},[55,2295,2296,2302,2308,2314,2320,2326],{},[58,2297,2298,2301],{},[61,2299,2300],{},"Web 版数据不本地","：隐私敏感选桌面或 Docker 自托管",[58,2303,2304,2307],{},[61,2305,2306],{},"国内连海外模型走中转","：直连 OpenAI \u002F Claude 不稳，配 OpenRouter \u002F Ofox \u002F 国内中转",[58,2309,2310,2313],{},[61,2311,2312],{},"Docker 自托管暴露公网","：上反代 + HTTPS + Auth + 备份数据库",[58,2315,2316,2319],{},[61,2317,2318],{},"嵌入模型中文优化","：默认嵌入对中文一般，配 bge-m3 \u002F 硅基流动 Pro 版",[58,2321,2322,2325],{},[61,2323,2324],{},"插件市场审一遍","：第三方插件可执行代码，团队部署谨慎启用",[58,2327,2328,2331],{},[61,2329,2330],{},"同步选 Cloud vs WebDAV","：团队多端走 LobeHub Cloud；个人多设备 WebDAV 即可",[35,2333,416],{"id":415},[55,2335,2336,2339,2342,2345,2348,2351,2354,2357],{},[58,2337,2338],{},"✅ Web + 桌面双形态需求",[58,2340,2341],{},"✅ Docker 自托管 \u002F 团队共享",[58,2343,2344],{},"✅ 多模型对比 \u002F 选型",[58,2346,2347],{},"✅ 中文重度用户",[58,2349,2350],{},"✅ 助手市场 \u002F 插件生态用户",[58,2352,2353],{},"❌ 强企业 RBAC + 多租户（Open WebUI 更完善）",[58,2355,2356],{},"❌ 只要桌面 + 数据完全本地（Cherry Studio 同样优秀）",[58,2358,2359],{},"❌ 完全不会碰 Docker",[35,2361,442],{"id":442},[55,2363,2364,2368,2372,2376],{},[58,2365,2366],{},[448,2367,463],{"href":462},[58,2369,2370],{},[448,2371,457],{"href":456},[58,2373,2374],{},[448,2375,953],{"href":537},[58,2377,2378],{},[448,2379,469],{"href":468},[35,2381,472],{"id":472},[202,2383,2384,2391,2398],{},[58,2385,2386,2387],{},"LobeChat GitHub 仓库（72k+ stars，MIT）",[448,2388,2389],{"href":2389,"rel":2390},"https:\u002F\u002Fgithub.com\u002Flobehub\u002Flobe-chat",[482],[58,2392,2393,2394],{},"腾讯云开发者社区 — Lobe Chat 本地化 AI 聊天终极桌面客户端（2026-01）",[448,2395,2396],{"href":2396,"rel":2397},"https:\u002F\u002Fcloud.tencent.com\u002Fdeveloper\u002Farticle\u002F2622150",[482],[58,2399,2400,2401],{},"Ofox.ai — LobeChat 完全配置指南 2026（2026-04-17）",[448,2402,2403],{"href":2403,"rel":2404},"https:\u002F\u002Fofox.ai\u002Fzh\u002Fblog\u002Flobechat-api-configuration-guide-2026",[482],{"title":498,"searchDepth":499,"depth":499,"links":2406},[2407,2408,2409,2410,2411,2412,2413,2414,2415,2416],{"id":37,"depth":502,"text":38},{"id":53,"depth":502,"text":53},{"id":129,"depth":502,"text":129},{"id":2082,"depth":502,"text":2083},{"id":200,"depth":502,"text":200},{"id":238,"depth":502,"text":238},{"id":380,"depth":502,"text":380},{"id":415,"depth":502,"text":416},{"id":442,"depth":502,"text":442},{"id":472,"depth":502,"text":472},"\u002Fimg\u002Ftools\u002Flobe-chat.webp","LobeChat 真实评测：LobeHub 团队开源 AI 聊天框架，GitHub 72k+ stars、MIT 协议。Web + 桌面（Win\u002FMac\u002FLinux\u002FDocker）双形态，支持 OpenAI \u002F Claude \u002F Gemini \u002F DeepSeek \u002F Qwen \u002F Ollama 等 80+ 模型，内置 RAG 知识库 + 插件市场 + 助手市场 + 多模型对比。",[2420,2423,2426,2429],{"q":2421,"a":2422},"Web 版 vs 桌面版 vs Docker 自托管，怎么选？","Web 版（chat.lobehub.com）最快上手但数据存 LobeHub 服务器；桌面版数据本地存、隐私好；Docker 自托管对团队 \u002F 公司部署最优，完全掌控数据。",{"q":2424,"a":2425},"支持哪些模型？","80+ 模型：OpenAI 全系列、Anthropic Claude、Google Gemini、DeepSeek、Qwen、Kimi、Moonshot、字节豆包、Groq、Together、OpenRouter、Ollama \u002F LM Studio 本地模型，以及任何 OpenAI 兼容 API。",{"q":2427,"a":2428},"多模型对比怎么用？","同一对话窗口里把消息广播给多个模型并排回答，选型 \u002F 评估特别有用——直接看 Claude 和 GPT 在同一 prompt 下的回答差异。",{"q":2430,"a":2431},"助手市场是什么？","LobeHub 维护的预设 AI 角色市场（代码审查 \u002F 翻译 \u002F 写作 \u002F 角色扮演等几百个），一键拉到本地用，省去自己写 System Prompt。",[1019,533],{},[2435,540,541,542,543],"web",[2437,2439],{"plan":2070,"price":547,"features":2438,"notes":1509},"全功能 \u002F 80+ 模型 \u002F 知识库 \u002F 插件 \u002F 助手市场",{"plan":2050,"price":2440,"features":2441,"notes":2442},"订阅制","云端托管 \u002F 免部署 \u002F 团队协作 \u002F 同步","chat.lobehub.com 注册即用","完全免费（MIT 开源） \u002F LobeHub Cloud 订阅",[553,554],{"power":557,"ux":557,"price":557,"cn_support":557,"stability":556},{"title":1247,"description":2418},"LobeChat - 开源 AI 聊天框架评测与部署 | AIHO",[2449,2451,2453],{"name":2450,"url":2389,"accessed":564},"LobeChat GitHub",{"name":2452,"url":2396,"accessed":564},"腾讯云开发者社区 — Lobe Chat 终极桌面客户端",{"name":2454,"url":2403,"accessed":564},"Ofox.ai — LobeChat 完全配置指南 2026","tools\u002Fcoding\u002Flocal\u002Flobe-chat","现代设计的开源 AI 聊天框架——Web + 桌面双形态、72k+ stars、多模型 + 知识库 + 插件市场",[512,2435,1973,1974,1531,2458,2459,579],"plugin","mcp","颜值与功能双优的多模型 AI 聊天客户端。要 Web + 桌面双形态、自托管 Docker、多模型对比、丰富助手市场——LobeChat 是综合最强；纯桌面体验 Cherry Studio 同样优秀。","https:\u002F\u002Flobehub.com","sAhQ5CHDobJa--z6NbtgysTQDwTBYsoGPXR0zat-5-w",1785660641129]