[{"data":1,"prerenderedAt":1945},["ShallowReactive",2],{"header-counts":3,"footer-counts":6,"alt-main-jan":8,"alt-list-jan":506},{"tools":4,"reviews":5},98,27,{"tools":4,"reviews":5,"playbooks":7,"news":5},23,{"id":9,"title":10,"alternatives":11,"api_compatible":15,"body":16,"category":470,"chinese_friendly":456,"cover":471,"description":472,"domestic":473,"extension":474,"faq":15,"free":475,"github":451,"languages":476,"lastVerified":478,"meta":479,"models":15,"navigation":475,"notSuitable":15,"opensource":475,"path":480,"pillar":481,"platforms":482,"priceTable":15,"pricing":486,"published":487,"relatedPlaybooks":15,"relatedReviews":15,"score":488,"self_host":473,"seo":491,"seoTitle":492,"slug":493,"sources":494,"stem":497,"suitable":15,"tagline":498,"tags":499,"updated":478,"verdict":504,"website":443,"__hash__":505},"tools\u002Ftools\u002Fcoding\u002Flocal\u002Fjan.md","Jan",[12,13,14],"coding\u002Flocal\u002Flm-studio","coding\u002Flocal\u002Fcherry-studio","coding\u002Flocal\u002Follama",null,{"type":17,"value":18,"toc":454},"minimark",[19,24,33,36,39,91,94,100,133,136,140,148,165,170,187,190,211,214,326,329,361,365,388,392,398,404,410,413,429,432,437],[20,21,23],"h2",{"id":22},"tldr","TL;DR",[25,26,27,28,32],"p",{},"Jan 是开源的本地 LLM 桌面客户端，定位是",[29,30,31],"strong",{},"ChatGPT 的离线替代品","。界面设计精美，操作体验接近 ChatGPT，支持 GGUF 模型一键下载、本地推理、多模型切换、插件扩展。AGPL 开源，完全免费。适合想要一个好看好用的本地 AI 聊天工具、隐私优先的用户。",[25,34,35],{},"适合：想要 ChatGPT 颜值和体验的本地替代、个人离线聊天、隐私敏感场景、非技术用户（GUI 友好）。不适合：需要 OpenAI 兼容 API 给应用接入（用 Ollama）、需要模型调参 \u002F 量化选择（用 LM Studio）、企业商用（AGPL 限制）。",[20,37,38],{"id":38},"核心能力",[40,41,42,49,55,61,67,73,79,85],"ul",{},[43,44,45,48],"li",{},[29,46,47],{},"ChatGPT 式界面","：聊天 UI 设计精美，多会话管理、Markdown 渲染、代码高亮",[43,50,51,54],{},[29,52,53],{},"一键下载模型","：内置模型市场，搜索 GGUF 模型点击下载，自动配置",[43,56,57,60],{},[29,58,59],{},"本地推理","：基于 llama.cpp，支持 CPU \u002F GPU 加速，完全离线运行",[43,62,63,66],{},[29,64,65],{},"多模型切换","：一个会话可切换不同模型对比输出，方便评估",[43,68,69,72],{},[29,70,71],{},"插件系统","：支持扩展功能，如网页搜索、文档分析、API 代理等",[43,74,75,78],{},[29,76,77],{},"远程 API 接入","：除了本地模型，也支持接 OpenAI \u002F Anthropic 等云端 API",[43,80,81,84],{},[29,82,83],{},"跨平台桌面 App","：Win \u002F Mac \u002F Linux 原生安装包，Electron 构建",[43,86,87,90],{},[29,88,89],{},"隐私优先","：所有数据本地存储，无遥测，无云端调用（本地模型模式）",[20,92,93],{"id":93},"价格",[95,96,97],"blockquote",{},[25,98,99],{},"以下信息为 2026-07-30 核实。",[101,102,103,118],"table",{},[104,105,106],"thead",{},[107,108,109,113,115],"tr",{},[110,111,112],"th",{},"方案",[110,114,93],{},[110,116,117],{},"说明",[119,120,121],"tbody",{},[107,122,123,127,130],{},[124,125,126],"td",{},"开源版",[124,128,129],{},"$0",[124,131,132],{},"完整功能，AGPL 协议",[25,134,135],{},"完全免费。注意 AGPL 协议：个人使用无限制，但二次开发 \u002F 商用需遵守开源传染条款。",[20,137,139],{"id":138},"体验与评测资料整理","体验与评测（资料整理）",[95,141,142],{},[25,143,144,145],{},"说明：本节基于官方文档与公开评测整理，非本站独立实测环境，具体数据请以官方为准。\n",[29,146,147],{},"亮点：",[40,149,150,153,156,159,162],{},[43,151,152],{},"界面设计是同类最佳：比 LM Studio \u002F GPT4All 好看很多，接近 ChatGPT 体验",[43,154,155],{},"模型下载体验顺滑：搜索 → 下载 → 使用，全程 GUI，零命令行",[43,157,158],{},"多模型对比实用：同一问题切换模型看不同回答，选模型很方便",[43,160,161],{},"插件系统有潜力：网页搜索插件让本地模型也能联网",[43,163,164],{},"支持云端 API 混用：本地模型 + GPT-4o 切换，一个客户端搞定",[25,166,167],{},[29,168,169],{},"踩坑：",[40,171,172,175,178,181,184],{},[43,173,174],{},"Electron 应用内存占用偏高，老设备偶有卡顿",[43,176,177],{},"模型管理不如 LM Studio：量化版本选择少，调参选项有限",[43,179,180],{},"API Server 功能弱：有 OpenAI 兼容端点但不如 Ollama 灵活",[43,182,183],{},"插件生态尚不成熟，可用插件不多",[43,185,186],{},"AGPL 协议对企业不友好，商用需注意合规",[20,188,189],{"id":189},"上手",[191,192,193,196,199,202,205,208],"ol",{},[43,194,195],{},"从 jan.ai 下载对应平台安装包",[43,197,198],{},"安装后打开 Jan，界面类似 ChatGPT",[43,200,201],{},"点击模型市场（Hub）→ 搜索推荐模型（Qwen2.5-7B \u002F Llama3.1-8B）",[43,203,204],{},"下载模型后，新建会话 → 选择模型 → 开始聊天",[43,206,207],{},"多模型对比：同一会话切换模型或开多个会话",[43,209,210],{},"接云端 API：Settings → API Keys → 填入 OpenAI Key 即可混用",[20,212,213],{"id":213},"对比",[101,215,216,234],{},[104,217,218],{},[107,219,220,223,225,228,231],{},[110,221,222],{},"维度",[110,224,10],{},[110,226,227],{},"LM Studio",[110,229,230],{},"Ollama",[110,232,233],{},"Cherry Studio",[119,235,236,252,266,281,295,309],{},[107,237,238,241,244,247,250],{},[124,239,240],{},"界面颜值",[124,242,243],{},"高",[124,245,246],{},"中",[124,248,249],{},"无 GUI",[124,251,243],{},[107,253,254,257,259,262,264],{},[124,255,256],{},"模型管理",[124,258,246],{},[124,260,261],{},"强",[124,263,261],{},[124,265,246],{},[107,267,268,271,274,277,279],{},[124,269,270],{},"API 接入",[124,272,273],{},"基础",[124,275,276],{},"✅",[124,278,261],{},[124,280,276],{},[107,282,283,286,288,291,293],{},[124,284,285],{},"插件扩展",[124,287,276],{},[124,289,290],{},"❌",[124,292,290],{},[124,294,276],{},[107,296,297,300,302,304,307],{},[124,298,299],{},"云端 API 混用",[124,301,276],{},[124,303,276],{},[124,305,306],{},"需配",[124,308,276],{},[107,310,311,314,317,320,323],{},[124,312,313],{},"开源协议",[124,315,316],{},"AGPL",[124,318,319],{},"闭源",[124,321,322],{},"MIT",[124,324,325],{},"Apache",[20,327,328],{"id":328},"避坑",[40,330,331,337,343,349,355],{},[43,332,333,336],{},[29,334,335],{},"别指望它做 API 服务器","：Jan 的 API Server 功能基础，给应用接入用 Ollama",[43,338,339,342],{},[29,340,341],{},"模型选对量化","：默认下载的可能不是最优量化，手动选 Q4_K_M 平衡速度质量",[43,344,345,348],{},[29,346,347],{},"Electron 吃内存","：8GB RAM 设备跑大模型 + Jan 本身会卡，关其他应用",[43,350,351,354],{},[29,352,353],{},"AGPL 商用注意","：企业内部署需法务确认 AGPL 合规",[43,356,357,360],{},[29,358,359],{},"插件别装太多","：部分插件质量参差，可能影响稳定性",[20,362,364],{"id":363},"适合-不适合","适合 \u002F 不适合",[40,366,367,370,373,376,379,382,385],{},[43,368,369],{},"✅ 想要 ChatGPT 颜值和体验的本地替代",[43,371,372],{},"✅ 个人离线聊天 \u002F 隐私优先场景",[43,374,375],{},"✅ 非技术用户（GUI 友好，零命令行）",[43,377,378],{},"✅ 本地 + 云端 API 混用需求",[43,380,381],{},"❌ 需要给应用 \u002F IDE 接入 API（用 Ollama）",[43,383,384],{},"❌ 需要精细模型调参 \u002F 量化管理（用 LM Studio）",[43,386,387],{},"❌ 企业商用（AGPL 限制）",[20,389,391],{"id":390},"faq","FAQ",[25,393,394,397],{},[29,395,396],{},"Q: Jan 和 LM Studio 怎么选？","\nA: 颜值和聊天体验选 Jan，模型管理和调参选 LM Studio。Jan 更像 ChatGPT，LM Studio 更像模型工具箱。两者都免费，可以都装。",[25,399,400,403],{},[29,401,402],{},"Q: 能给 Cursor \u002F Cline 接入吗？","\nA: Jan 有 OpenAI 兼容 API Server（默认端口 1337），理论上可以。但不如 Ollama 稳定灵活，推荐用 Ollama 做 API 服务器。",[25,405,406,409],{},[29,407,408],{},"Q: AGPL 协议影响个人使用吗？","\nA: 不影响。AGPL 主要约束网络服务分发场景。个人本地使用完全无限制。只有你把 Jan 改造后对外提供 SaaS 服务才需开源你的修改。",[20,411,412],{"id":412},"相关阅读",[25,414,415,420,421,420,425],{},[416,417,419],"a",{"href":418},"\u002Fcoding\u002Flocal\u002Fgpt4all.html","GPT4All"," · ",[416,422,424],{"href":423},"\u002Fcoding\u002Flocal\u002Fvllm.html","vLLM",[416,426,428],{"href":427},"\u002Fagent\u002Fdesktop\u002Fopen-interpreter.html","Open Interpreter",[20,430,431],{"id":431},"来源",[95,433,434],{},[25,435,436],{},"本文的价格、版本号与性能数据均参考以下官方渠道整理，可能随时间变动，请以官方实时信息为准。",[40,438,439,447],{},[43,440,441],{},[416,442,446],{"href":443,"rel":444},"https:\u002F\u002Fjan.ai",[445],"nofollow","官网",[43,448,449],{},[416,450,453],{"href":451,"rel":452},"https:\u002F\u002Fgithub.com\u002Fjanhq\u002Fjan",[445],"GitHub",{"title":455,"searchDepth":456,"depth":456,"links":457},"",3,[458,460,461,462,463,464,465,466,467,468,469],{"id":22,"depth":459,"text":23},2,{"id":38,"depth":459,"text":38},{"id":93,"depth":459,"text":93},{"id":138,"depth":459,"text":139},{"id":189,"depth":459,"text":189},{"id":213,"depth":459,"text":213},{"id":328,"depth":459,"text":328},{"id":363,"depth":459,"text":364},{"id":390,"depth":459,"text":391},{"id":412,"depth":459,"text":412},{"id":431,"depth":459,"text":431},"local","\u002Fimg\u002Ftools\u002Fjan.webp","Jan 真实评测：开源本地 LLM 桌面客户端（AGPL 协议），定位 ChatGPT 的离线替代，支持 GGUF 模型一键下载 + 本地推理 + 插件扩展。跨平台桌面 app，适合需要完全离线、隐私优先的本地 AI 聊天场景。",false,"md",true,[477],"en","2026-07-30",{},"\u002Ftools\u002Fcoding\u002Flocal\u002Fjan","coding",[483,484,485],"windows","macos","linux","Free \u002F 开源（AGPL）","2026-07-05",{"power":456,"ux":489,"price":490,"cn_support":456,"stability":456},4,5,{"title":10,"description":472},"Jan - 开源本地 LLM 桌面客户端评测 | AIHO","coding\u002Flocal\u002Fjan",[495,496],{"title":446,"url":443},{"title":453,"url":451},"tools\u002Fcoding\u002Flocal\u002Fjan","开源本地 LLM 桌面客户端，定位 ChatGPT 的离线替代",[470,500,501,502,503],"desktop","opensource","gguf","offline","颜值最高、最像 ChatGPT 的开源本地 LLM 客户端，离线聊天体验好；但 API 能力和模型管理不如 Ollama\u002FLM Studio，定位偏轻量个人使用。","A1THYvHqqGYW3sNLdKo_wmhGktyak4UnrJaiENKFhro",[507,1005,1465],{"id":508,"title":227,"alternatives":509,"api_compatible":15,"body":512,"category":470,"chinese_friendly":456,"cover":948,"description":949,"domestic":473,"extension":474,"faq":950,"free":475,"github":15,"languages":963,"lastVerified":965,"meta":966,"models":15,"navigation":475,"notSuitable":15,"opensource":473,"path":967,"pillar":481,"platforms":968,"priceTable":969,"pricing":978,"published":979,"relatedPlaybooks":980,"relatedReviews":15,"score":983,"self_host":475,"seo":984,"seoTitle":985,"slug":12,"sources":986,"stem":994,"suitable":15,"tagline":995,"tags":996,"updated":989,"verdict":1002,"website":1003,"__hash__":1004},"tools\u002Ftools\u002Fcoding\u002Flocal\u002Flm-studio.md",[14,510,13,511],"coding\u002Flocal\u002Fopen-webui","coding\u002Flocal\u002Flobe-chat",{"type":17,"value":513,"toc":936},[514,516,524,527,529,586,588,602,607,611,615,632,636,653,655,681,683,824,826,858,860,883,885,911,913],[20,515,23],{"id":22},[25,517,518,519,523],{},"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 + ",[520,521,522],"code",{},"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。",[25,525,526],{},"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）。",[20,528,38],{"id":38},[40,530,531,537,543,549,559,568,574,580],{},[43,532,533,536],{},[29,534,535],{},"模型浏览器","：内置 Hugging Face 检索，按 GGUF \u002F MLX \u002F 大小筛选、一键下载",[43,538,539,542],{},[29,540,541],{},"聊天界面","：System Prompt \u002F temperature \u002F top-p \u002F context size 可视化调参",[43,544,545,548],{},[29,546,547],{},"多模型并存 \u002F 切换","：同时加载多模型在不同会话中比较",[43,550,551,554,555,558],{},[29,552,553],{},"OpenAI 兼容 Local Server","：",[520,556,557],{},"http:\u002F\u002Flocalhost:1234\u002Fv1","，任何 SDK 即接即用",[43,560,561,554,564,567],{},[29,562,563],{},"Headless \u002F CLI",[520,565,566],{},"lms server start --port 1234","，无 GUI 可跑",[43,569,570,573],{},[29,571,572],{},"PDF \u002F 文档对话","：内置基础 RAG，丢文件就能聊",[43,575,576,579],{},[29,577,578],{},"MLX 原生支持（Mac）","：M1+ 上比 GGUF + Metal 快 30–50%",[43,581,582,585],{},[29,583,584],{},"持续批处理","：Codersera 2026 测得 50–90 tok\u002Fs（消费级 GPU + 中等模型）",[20,587,93],{"id":93},[40,589,590,596],{},[43,591,592,595],{},[29,593,594],{},"个人 \u002F 评估","：免费，全功能可用",[43,597,598,601],{},[29,599,600],{},"商用","：邮件 \u002F 官网联系 LM Studio 团队",[95,603,604],{},[25,605,606],{},"模型本身免费（开源权重），LM Studio 不抽水任何 token 费用。",[20,608,610],{"id":609},"实测mac-m2-pro-qwen3-coder-7b-gguf-q4_k_m","实测（Mac M2 Pro + Qwen3-Coder-7B GGUF Q4_K_M）",[25,612,613],{},[29,614,147],{},[40,616,617,620,623,626,629],{},[43,618,619],{},"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",[43,621,622],{},"加载 7B Q4 模型 \u003C 3 秒，生成 ~75 tok\u002Fs",[43,624,625],{},"Local Server 开了 Cursor 直接接 baseURL → 本地代码补全零成本",[43,627,628],{},"MLX 版同模型 ~110 tok\u002Fs，差距显著",[43,630,631],{},"多窗口加载 2 个模型并排测，调 prompt 直观",[25,633,634],{},[29,635,169],{},[40,637,638,641,644,647,650],{},[43,639,640],{},"模型库依赖 Hugging Face，国内访问要镜像 \u002F 代理",[43,642,643],{},"GPU 显存吃满后会自动 offload 到 CPU，无提示就慢下来",[43,645,646],{},"Headless 模式相对 Ollama 偏新，文档稍少",[43,648,649],{},"闭源应用（虽免费），不适合企业合规挂钩",[43,651,652],{},"中文 UI 可用但部分菜单仍英文",[20,654,189],{"id":189},[191,656,657,660,663,666,669,676],{},[43,658,659],{},"lmstudio.ai 下载（Mac \u002F Windows \u002F Linux）",[43,661,662],{},"打开 → Discover 标签 → 搜模型（如 qwen3-coder、deepseek-v3 GGUF\u002FMLX）→ Download",[43,664,665],{},"Chat 标签 → 选模型 → 调参聊天",[43,667,668],{},"Local Server 标签 → Start Server → 默认端口 1234",[43,670,671,672,675],{},"在你的应用里：",[520,673,674],{},"baseURL = \"http:\u002F\u002Flocalhost:1234\u002Fv1\"","，API Key 任意",[43,677,678,679],{},"Headless：",[520,680,566],{},[20,682,213],{"id":213},[101,684,685,701],{},[104,686,687],{},[107,688,689,691,693,695,698],{},[110,690,222],{},[110,692,227],{},[110,694,230],{},[110,696,697],{},"Open WebUI",[110,699,700],{},"llama.cpp",[119,702,703,720,737,751,766,781,795,809],{},[107,704,705,708,711,714,717],{},[124,706,707],{},"形态",[124,709,710],{},"GUI + CLI",[124,712,713],{},"CLI Daemon",[124,715,716],{},"Docker UI",[124,718,719],{},"二进制",[107,721,722,725,728,731,734],{},[124,723,724],{},"模型浏览",[124,726,727],{},"✅ 内置",[124,729,730],{},"CLI pull",[124,732,733],{},"无",[124,735,736],{},"手动",[107,738,739,742,744,746,749],{},[124,740,741],{},"参数调优 GUI",[124,743,276],{},[124,745,290],{},[124,747,748],{},"部分",[124,750,290],{},[107,752,753,756,759,762,764],{},[124,754,755],{},"OpenAI 兼容 API",[124,757,758],{},"✅ :1234",[124,760,761],{},"✅ :11434",[124,763,276],{},[124,765,276],{},[107,767,768,771,773,776,779],{},[124,769,770],{},"MLX (Mac)",[124,772,276],{},[124,774,775],{},"✅ 0.19+",[124,777,778],{},"–",[124,780,778],{},[107,782,783,786,789,791,793],{},[124,784,785],{},"多用户并发",[124,787,788],{},"弱",[124,790,788],{},[124,792,276],{},[124,794,246],{},[107,796,797,800,803,805,807],{},[124,798,799],{},"开源",[124,801,802],{},"闭源（免费）",[124,804,322],{},[124,806,322],{},[124,808,322],{},[107,810,811,814,817,820,822],{},[124,812,813],{},"上手难度",[124,815,816],{},"极低",[124,818,819],{},"低",[124,821,246],{},[124,823,243],{},[20,825,328],{"id":328},[40,827,828,834,840,846,852],{},[43,829,830,833],{},[29,831,832],{},"国内下模型走镜像","：HF 直连慢 \u002F 卡，配 HF_ENDPOINT=hf-mirror.com",[43,835,836,839],{},[29,837,838],{},"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",[43,841,842,845],{},[29,843,844],{},"MLX 优先（Mac M 系列）","：能下 MLX 版就别下 GGUF，速度差距明显",[43,847,848,851],{},[29,849,850],{},"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",[43,853,854,857],{},[29,855,856],{},"闭源合规要核","：企业内部使用前查 license；商用必须联系官方",[20,859,364],{"id":363},[40,861,862,865,868,871,874,877,880],{},[43,863,864],{},"✅ 本地 LLM 入门 \u002F 评估",[43,866,867],{},"✅ Mac M 系列用户",[43,869,870],{},"✅ 想给 Cursor \u002F Cline 接本地 OpenAI 兼容 endpoint",[43,872,873],{},"✅ 需要 GUI 调参 \u002F 模型比较",[43,875,876],{},"❌ 多用户并发生产服务",[43,878,879],{},"❌ 嵌入式 \u002F 边缘设备",[43,881,882],{},"❌ 强合规 \u002F 必须开源审计",[20,884,412],{"id":412},[40,886,887,893,899,905],{},[43,888,889],{},[416,890,892],{"href":891},"\u002Ftools\u002Fcoding\u002Flocal\u002Follama","Ollama 评测",[43,894,895],{},[416,896,898],{"href":897},"\u002Ftools\u002Fcoding\u002Flocal\u002Fopen-webui","Open WebUI 评测",[43,900,901],{},[416,902,904],{"href":903},"\u002Ftools\u002Fcoding\u002Flocal\u002Fcherry-studio","Cherry Studio 评测",[43,906,907],{},[416,908,910],{"href":909},"\u002Fplaybook\u002Fonboarding\u002Fclaude-code-getting-started","Claude Code 上手 Playbook",[20,912,431],{"id":431},[191,914,915,922,929],{},[43,916,917,918],{},"LM Studio 官网 ",[416,919,920],{"href":920,"rel":921},"https:\u002F\u002Flmstudio.ai\u002F",[445],[43,923,924,925],{},"Codersera — LM Studio Complete Guide 2026 ",[416,926,927],{"href":927,"rel":928},"https:\u002F\u002Fcodersera.com\u002Fblog\u002Flm-studio-complete-guide-2026\u002F",[445],[43,930,931,932],{},"Codersera — Ollama vs LM Studio vs vLLM vs llama.cpp vs MLX 2026 ",[416,933,934],{"href":934,"rel":935},"https:\u002F\u002Fcodersera.com\u002Fblog\u002Follama-vs-lm-studio-vs-vllm-vs-llama-cpp-vs-mlx-2026\u002F",[445],{"title":455,"searchDepth":456,"depth":456,"links":937},[938,939,940,941,942,943,944,945,946,947],{"id":22,"depth":459,"text":23},{"id":38,"depth":459,"text":38},{"id":93,"depth":459,"text":93},{"id":609,"depth":459,"text":610},{"id":189,"depth":459,"text":189},{"id":213,"depth":459,"text":213},{"id":328,"depth":459,"text":328},{"id":363,"depth":459,"text":364},{"id":412,"depth":459,"text":412},{"id":431,"depth":459,"text":431},"\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。",[951,954,957,960],{"q":952,"a":953},"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":955,"a":956},"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":958,"a":959},"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":961,"a":962},"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。",[477,964],"zh","2026-08-02",{},"\u002Ftools\u002Fcoding\u002Flocal\u002Flm-studio",[483,484,485],[970,974],{"plan":594,"price":971,"features":972,"notes":973},"免费","全功能 GUI + Headless API + GGUF\u002FMLX","供个人 \u002F 评估使用",{"plan":600,"price":975,"features":976,"notes":977},"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","2026-06-19",[981,982],"onboarding\u002Frag-pipeline-build","onboarding\u002Fclaude-code-getting-started",{"power":489,"ux":490,"price":490,"cn_support":456,"stability":489},{"title":227,"description":949},"LM Studio 评测 2026：本地运行开源大模型，图形化界面，AI 模型管理",[987,990,992],{"name":988,"url":920,"accessed":989},"LM Studio 官网","2026-06-24",{"name":991,"url":927,"accessed":989},"Codersera — LM Studio Complete Guide 2026",{"name":993,"url":934,"accessed":989},"Codersera — Ollama vs LM Studio vs vLLM 2026","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",[470,997,502,998,999,1000,1001],"gui","mlx","llama-cpp","mac","openai-compatible","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":1006,"title":233,"alternatives":1007,"api_compatible":15,"body":1008,"category":470,"chinese_friendly":490,"cover":1413,"description":1414,"domestic":475,"extension":474,"faq":1415,"free":475,"github":1428,"languages":1429,"lastVerified":965,"meta":1430,"models":15,"navigation":475,"notSuitable":15,"opensource":475,"path":903,"pillar":481,"platforms":1431,"priceTable":1433,"pricing":1441,"published":979,"relatedPlaybooks":1442,"relatedReviews":15,"score":1444,"self_host":475,"seo":1445,"seoTitle":1446,"slug":13,"sources":1447,"stem":1454,"suitable":15,"tagline":1455,"tags":1456,"updated":989,"verdict":1462,"website":1463,"__hash__":1464},"tools\u002Ftools\u002Fcoding\u002Flocal\u002Fcherry-studio.md",[511,12,14,510],{"type":17,"value":1009,"toc":1401},[1010,1012,1015,1018,1020,1064,1066,1079,1084,1088,1092,1109,1113,1130,1132,1156,1158,1292,1294,1326,1328,1351,1353,1376,1378],[20,1011,23],{"id":22},[25,1013,1014],{},"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。",[25,1016,1017],{},"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。",[20,1019,38],{"id":38},[40,1021,1022,1028,1034,1040,1046,1052,1058],{},[43,1023,1024,1027],{},[29,1025,1026],{},"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",[43,1029,1030,1033],{},[29,1031,1032],{},"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",[43,1035,1036,1039],{},[29,1037,1038],{},"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",[43,1041,1042,1045],{},[29,1043,1044],{},"MCP 协议","：扩展工具调用 \u002F 联网搜索 \u002F 文件操作",[43,1047,1048,1051],{},[29,1049,1050],{},"数据本地优先","：对话历史本地存储，WebDAV 同步，不上传第三方",[43,1053,1054,1057],{},[29,1055,1056],{},"多模态","：图片识别 \u002F PDF 阅读 \u002F Markdown + Mermaid + 代码高亮",[43,1059,1060,1063],{},[29,1061,1062],{},"AI 绘画 + 翻译","：内置主流 SD \u002F DALL·E \u002F 翻译 API 集成",[20,1065,93],{"id":93},[40,1067,1068,1073],{},[43,1069,1070,1072],{},[29,1071,126],{},"：完全免费，AGPL-3.0",[43,1074,1075,1078],{},[29,1076,1077],{},"Enterprise","：私有化部署 + 团队协作 + 资源管控，联系销售",[95,1080,1081],{},[25,1082,1083],{},"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。",[20,1085,1087],{"id":1086},"实测mac-m2-中型知识库","实测（Mac M2 + 中型知识库）",[25,1089,1090],{},[29,1091,147],{},[40,1093,1094,1097,1100,1103,1106],{},[43,1095,1096],{},"中文 UI \u002F 文档 \u002F 社区都顶级，零门槛上手",[43,1098,1099],{},"本地 RAG 拖入 30+ PDF 后向量化 \u003C 2 分钟（用 bge-m3）",[43,1101,1102],{},"多模型并排回答：让 Claude \u002F GPT \u002F DeepSeek 同回一个问题做比较",[43,1104,1105],{},"MCP 接 Brave Search + 自定义工具流畅",[43,1107,1108],{},"WebDAV 同步坚果云 \u002F 阿里云盘，桌面 + 移动设备数据互通",[25,1110,1111],{},[29,1112,169],{},[40,1114,1115,1118,1121,1124,1127],{},[43,1116,1117],{},"没有 Web 端 \u002F Docker 自托管（要这个用 LobeChat）",[43,1119,1120],{},"iOS 版尚未发布（roadmap 中）",[43,1122,1123],{},"大型 PDF（>100 MB）向量化偶有失败，要切小",[43,1125,1126],{},"助手市场质量参差，要自筛",[43,1128,1129],{},"模型 API 调用全靠你自己付费，新手要先理解 API Key 概念",[20,1131,189],{"id":189},[191,1133,1134,1137,1140,1147,1150,1153],{},[43,1135,1136],{},"cherry-ai.com 下载客户端（或 GitHub releases）",[43,1138,1139],{},"设置 → 模型服务 → 填 OpenAI \u002F Claude \u002F DeepSeek API Key",[43,1141,1142,1143],{},"（可选）本地：装 Ollama → Cherry Studio 自动识别 endpoint ",[416,1144,1145],{"href":1145,"rel":1146},"http:\u002F\u002Flocalhost:11434",[445],[43,1148,1149],{},"新建知识库 → 拖文件 \u002F 加网址 → 等向量化",[43,1151,1152],{},"新对话 → 选模型 → 勾知识库 → 提问",[43,1154,1155],{},"进阶：自定义助手（System Prompt）+ MCP 扩展工具",[20,1157,213],{"id":213},[101,1159,1160,1175],{},[104,1161,1162],{},[107,1163,1164,1166,1168,1171,1173],{},[110,1165,222],{},[110,1167,233],{},[110,1169,1170],{},"LobeChat",[110,1172,227],{},[110,1174,697],{},[119,1176,1177,1192,1206,1220,1233,1248,1263,1276],{},[107,1178,1179,1181,1184,1187,1189],{},[124,1180,707],{},[124,1182,1183],{},"桌面",[124,1185,1186],{},"Web + 桌面",[124,1188,1183],{},[124,1190,1191],{},"Docker \u002F 桌面",[107,1193,1194,1196,1199,1201,1204],{},[124,1195,1026],{},[124,1197,1198],{},"✅ 云 + 本地",[124,1200,1198],{},[124,1202,1203],{},"本地为主",[124,1205,1198],{},[107,1207,1208,1211,1214,1216,1218],{},[124,1209,1210],{},"知识库 RAG",[124,1212,1213],{},"✅ 强",[124,1215,1213],{},[124,1217,788],{},[124,1219,276],{},[107,1221,1222,1225,1227,1229,1231],{},[124,1223,1224],{},"MCP",[124,1226,276],{},[124,1228,276],{},[124,1230,788],{},[124,1232,276],{},[107,1234,1235,1238,1241,1244,1246],{},[124,1236,1237],{},"自托管 \u002F Web",[124,1239,1240],{},"无 Web",[124,1242,1243],{},"✅ Docker",[124,1245,733],{},[124,1247,1243],{},[107,1249,1250,1253,1256,1258,1261],{},[124,1251,1252],{},"中文",[124,1254,1255],{},"5\u002F5",[124,1257,1255],{},[124,1259,1260],{},"4\u002F5",[124,1262,1260],{},[107,1264,1265,1267,1270,1272,1274],{},[124,1266,313],{},[124,1268,1269],{},"AGPL-3.0",[124,1271,322],{},[124,1273,802],{},[124,1275,322],{},[107,1277,1278,1281,1284,1287,1289],{},[124,1279,1280],{},"GitHub Stars",[124,1282,1283],{},"60k+",[124,1285,1286],{},"72k+",[124,1288,778],{},[124,1290,1291],{},"126k+",[20,1293,328],{"id":328},[40,1295,1296,1302,1308,1314,1320],{},[43,1297,1298,1301],{},[29,1299,1300],{},"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",[43,1303,1304,1307],{},[29,1305,1306],{},"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",[43,1309,1310,1313],{},[29,1311,1312],{},"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",[43,1315,1316,1319],{},[29,1317,1318],{},"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",[43,1321,1322,1325],{},[29,1323,1324],{},"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",[20,1327,364],{"id":363},[40,1329,1330,1333,1336,1339,1342,1345,1348],{},[43,1331,1332],{},"✅ 中文用户、AI 重度使用 \u002F 多模型管理",[43,1334,1335],{},"✅ 需要本地 RAG 知识库",[43,1337,1338],{},"✅ 关注数据隐私 \u002F 本地存储",[43,1340,1341],{},"✅ 想用 Ollama \u002F LM Studio 本地模型",[43,1343,1344],{},"❌ 需要 Web 端 \u002F Docker 自托管",[43,1346,1347],{},"❌ 团队多人共享 \u002F SSO",[43,1349,1350],{},"❌ iOS 主力用户",[20,1352,412],{"id":412},[40,1354,1355,1361,1366,1370],{},[43,1356,1357],{},[416,1358,1360],{"href":1359},"\u002Ftools\u002Fcoding\u002Flocal\u002Flobe-chat","LobeChat 评测",[43,1362,1363],{},[416,1364,1365],{"href":967},"LM Studio 评测",[43,1367,1368],{},[416,1369,892],{"href":891},[43,1371,1372],{},[416,1373,1375],{"href":1374},"\u002Fplaybook\u002Fonboarding\u002Frag-pipeline-build","RAG Pipeline 搭建 Playbook",[20,1377,431],{"id":431},[191,1379,1380,1387,1394],{},[43,1381,1382,1383],{},"Cherry Studio 官网（功能 + 下载）",[416,1384,1385],{"href":1385,"rel":1386},"https:\u002F\u002Fwww.cherry-ai.com\u002F",[445],[43,1388,1389,1390],{},"MBLUO Studio — Cherry Studio 评测 2026 ",[416,1391,1392],{"href":1392,"rel":1393},"https:\u002F\u002Fmbluostudio.com\u002Ftools\u002Fcherry-studio",[445],[43,1395,1396,1397],{},"Cursor IDE 博客 — Cherry Studio 完全指南（2025-03）",[416,1398,1399],{"href":1399,"rel":1400},"https:\u002F\u002Fwww.cursor-ide.com\u002Fblog\u002Fcherry-studio-guide",[445],{"title":455,"searchDepth":456,"depth":456,"links":1402},[1403,1404,1405,1406,1407,1408,1409,1410,1411,1412],{"id":22,"depth":459,"text":23},{"id":38,"depth":459,"text":38},{"id":93,"depth":459,"text":93},{"id":1086,"depth":459,"text":1087},{"id":189,"depth":459,"text":189},{"id":213,"depth":459,"text":213},{"id":328,"depth":459,"text":328},{"id":363,"depth":459,"text":364},{"id":412,"depth":459,"text":412},{"id":431,"depth":459,"text":431},"\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。",[1416,1419,1422,1425],{"q":1417,"a":1418},"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":1420,"a":1421},"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":1423,"a":1424},"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":1426,"a":1427},"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",[964,477],{},[483,484,485,1432],"android",[1434,1437],{"plan":126,"price":971,"features":1435,"notes":1436},"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":1077,"price":1438,"features":1439,"notes":1440},"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",[981,1443],"onboarding\u002Fcursor-mcp-deep-integration",{"power":489,"ux":490,"price":490,"cn_support":490,"stability":489},{"title":233,"description":1414},"Cherry Studio 评测 2026：AI 客户端工具，多模型桌面助手，开源免费",[1448,1450,1452],{"name":1449,"url":1385,"accessed":989},"Cherry Studio 官网",{"name":1451,"url":1392,"accessed":989},"MBLUO Studio — Cherry Studio 评测",{"name":1453,"url":1399,"accessed":989},"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",[470,500,1457,1458,1459,1460,1461],"multi-model","knowledge-base","rag","open-source","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":1466,"title":230,"alternatives":1467,"api_compatible":1468,"body":1483,"category":470,"chinese_friendly":456,"cover":1900,"description":1901,"domestic":473,"extension":474,"faq":1902,"free":475,"github":1915,"languages":1916,"lastVerified":965,"meta":1917,"models":15,"navigation":475,"notSuitable":15,"opensource":475,"path":891,"pillar":481,"platforms":1918,"priceTable":1920,"pricing":1924,"published":979,"relatedPlaybooks":1925,"relatedReviews":15,"score":1926,"self_host":475,"seo":1927,"seoTitle":1928,"slug":14,"sources":1929,"stem":1935,"suitable":15,"tagline":1936,"tags":1937,"updated":989,"verdict":1942,"website":1943,"__hash__":1944},"tools\u002Ftools\u002Fcoding\u002Flocal\u002Follama.md",[12,510,13,511],[1469,1470,1471,1472,1473,1474,1475,1476,1477,1478,1479,1480,1481,230,1482],"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":17,"value":1484,"toc":1888},[1485,1487,1494,1497,1499,1567,1569,1572,1576,1580,1600,1604,1635,1637,1671,1673,1786,1788,1820,1822,1845,1847,1865,1867],[20,1486,23],{"id":22},[25,1488,1489,1490,1493],{},"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，",[520,1491,1492],{},"ollama pull"," 一键拉。",[25,1495,1496],{},"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）。",[20,1498,38],{"id":38},[40,1500,1501,1507,1515,1521,1528,1543,1549,1555,1561],{},[43,1502,1503,1506],{},[29,1504,1505],{},"后台 Daemon","：开机自启，应用调用零延迟",[43,1508,1509,554,1512],{},[29,1510,1511],{},"CLI",[520,1513,1514],{},"ollama pull \u002F run \u002F list \u002F show \u002F create \u002F serve",[43,1516,1517,1520],{},[29,1518,1519],{},"Modelfile","：类 Dockerfile 注册任意 GGUF，配 SYSTEM \u002F PARAMETER \u002F TEMPLATE",[43,1522,1523,554,1525],{},[29,1524,755],{},[520,1526,1527],{},"http:\u002F\u002Flocalhost:11434\u002Fv1\u002Fchat\u002Fcompletions",[43,1529,1530,554,1533,1536,1537,1536,1540],{},[29,1531,1532],{},"原生 API",[520,1534,1535],{},"\u002Fapi\u002Fchat","、",[520,1538,1539],{},"\u002Fapi\u002Fgenerate",[520,1541,1542],{},"\u002Fapi\u002Fembeddings",[43,1544,1545,1548],{},[29,1546,1547],{},"模型库","：官方注册表内置 Llama \u002F Qwen \u002F DeepSeek \u002F Gemma \u002F Mistral \u002F GPT-OSS 等",[43,1550,1551,1554],{},[29,1552,1553],{},"MLX 加速（Mac）","：0.19+ 起 M 系列自动用 MLX",[43,1556,1557,1560],{},[29,1558,1559],{},"量化","：默认 Q4_K_M、支持 Q5 \u002F Q8 \u002F FP16",[43,1562,1563,1566],{},[29,1564,1565],{},"跨平台","：Win \u002F Mac \u002F Linux 安装包，Docker 官方镜像",[20,1568,93],{"id":93},[25,1570,1571],{},"完全免费、MIT 开源、商用免费。",[20,1573,1575],{"id":1574},"实测m2-pro-qwen3-coder-7b-q4","实测（M2 Pro + Qwen3-Coder-7B Q4）",[25,1577,1578],{},[29,1579,147],{},[40,1581,1582,1588,1591,1594,1597],{},[43,1583,1584,1587],{},[520,1585,1586],{},"ollama run qwen3-coder:7b"," 一行起飞，3 秒进交互",[43,1589,1590],{},"REST API 配 Cursor \u002F Cline \u002F Continue 几乎全工具开箱即用",[43,1592,1593],{},"Modelfile 写自定义编码助手（low temperature + system prompt + 16K context）几分钟搞定",[43,1595,1596],{},"多模型并存，按需切换，内存占用合理",[43,1598,1599],{},"Mac M 系列 MLX 后比旧 GGUF 模式快显著",[25,1601,1602],{},[29,1603,169],{},[40,1605,1606,1616,1622,1629,1632],{},[43,1607,1608,1609,1612,1613],{},"默认 ",[520,1610,1611],{},"num_ctx"," 偏小（2048），跑长上下文要在 Modelfile 加 ",[520,1614,1615],{},"PARAMETER num_ctx 16384",[43,1617,1618,1619],{},"模型默认走 0.0.0.0:11434 ↔ Docker 容器互访要 ",[520,1620,1621],{},"--add-host=host.docker.internal:host-gateway",[43,1623,1624,1625,1628],{},"国内 ",[520,1626,1627],{},"ollama.com\u002Flibrary"," 下载偶有慢，可手动 HF 下 GGUF + Modelfile 自建",[43,1630,1631],{},"多用户并发吞吐显著低于 vLLM",[43,1633,1634],{},"没有 GUI，模型浏览 \u002F 参数面板要走 LM Studio \u002F Open WebUI 配合",[20,1636,189],{"id":189},[191,1638,1639,1645,1651,1656,1662,1668],{},[43,1640,1641,1644],{},[520,1642,1643],{},"curl -fsSL https:\u002F\u002Follama.ai\u002Finstall.sh | sh","（Mac \u002F Linux）；Windows winget",[43,1646,1647,1650],{},[520,1648,1649],{},"ollama pull qwen3-coder:7b","（按需换模型）",[43,1652,1653,1655],{},[520,1654,1586],{}," 直接聊",[43,1657,1658,1659],{},"应用接入：baseURL = ",[520,1660,1661],{},"http:\u002F\u002Flocalhost:11434\u002Fv1",[43,1663,1664,1665],{},"自定义：写 Modelfile → ",[520,1666,1667],{},"ollama create my-coder -f Modelfile",[43,1669,1670],{},"进阶：装 Open WebUI 做前端 \u002F 多人共享",[20,1672,213],{"id":213},[101,1674,1675,1689],{},[104,1676,1677],{},[107,1678,1679,1681,1683,1685,1687],{},[110,1680,222],{},[110,1682,230],{},[110,1684,227],{},[110,1686,424],{},[110,1688,700],{},[119,1690,1691,1707,1719,1732,1745,1761,1773],{},[107,1692,1693,1695,1698,1701,1704],{},[124,1694,707],{},[124,1696,1697],{},"CLI + Daemon",[124,1699,1700],{},"GUI + Headless",[124,1702,1703],{},"Python Server",[124,1705,1706],{},"C++ 二进制",[107,1708,1709,1711,1713,1715,1717],{},[124,1710,189],{},[124,1712,816],{},[124,1714,816],{},[124,1716,246],{},[124,1718,243],{},[107,1720,1721,1723,1725,1728,1730],{},[124,1722,724],{},[124,1724,1511],{},[124,1726,1727],{},"✅ GUI",[124,1729,733],{},[124,1731,733],{},[107,1733,1734,1737,1739,1741,1743],{},[124,1735,1736],{},"OpenAI 兼容",[124,1738,761],{},[124,1740,758],{},[124,1742,276],{},[124,1744,276],{},[107,1746,1747,1750,1753,1756,1759],{},[124,1748,1749],{},"多用户吞吐",[124,1751,1752],{},"弱（~40 tok\u002Fs）",[124,1754,1755],{},"中（50–90）",[124,1757,1758],{},"强（800–12500）",[124,1760,246],{},[107,1762,1763,1765,1767,1769,1771],{},[124,1764,770],{},[124,1766,775],{},[124,1768,276],{},[124,1770,748],{},[124,1772,778],{},[107,1774,1775,1777,1779,1781,1784],{},[124,1776,799],{},[124,1778,322],{},[124,1780,319],{},[124,1782,1783],{},"Apache 2.0",[124,1785,322],{},[20,1787,328],{"id":328},[40,1789,1790,1796,1802,1808,1814],{},[43,1791,1792,1795],{},[29,1793,1794],{},"num_ctx 一定要设","：默认 2K 太小，跑代码 \u002F 长文档要 16K+",[43,1797,1798,1801],{},[29,1799,1800],{},"Modelfile 模板别漏 TEMPLATE","：错的 chat template 会让模型输出乱码 \u002F 不停",[43,1803,1804,1807],{},[29,1805,1806],{},"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×",[43,1809,1810,1813],{},[29,1811,1812],{},"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",[43,1815,1816,1819],{},[29,1817,1818],{},"Mac 让它自动用 MLX","：升 0.19+；不要手动强制 GGUF + Metal",[20,1821,364],{"id":363},[40,1823,1824,1827,1830,1833,1836,1839,1842],{},[43,1825,1826],{},"✅ 应用 \u002F IDE 接本地模型（Cursor \u002F Cline \u002F Continue）",[43,1828,1829],{},"✅ 个人 \u002F 评估 \u002F 脚本自动化",[43,1831,1832],{},"✅ Modelfile 自定义系统 prompt + 参数",[43,1834,1835],{},"✅ Mac M 系列 MLX 用户",[43,1837,1838],{},"❌ 多用户并发生产服务（用 vLLM）",[43,1840,1841],{},"❌ GUI 调参 \u002F 模型浏览（配 LM Studio \u002F Open WebUI）",[43,1843,1844],{},"❌ 极致单卡吞吐研究（直接 llama.cpp \u002F vLLM）",[20,1846,412],{"id":412},[40,1848,1849,1853,1857,1861],{},[43,1850,1851],{},[416,1852,1365],{"href":967},[43,1854,1855],{},[416,1856,898],{"href":897},[43,1858,1859],{},[416,1860,904],{"href":903},[43,1862,1863],{},[416,1864,1375],{"href":1374},[20,1866,431],{"id":431},[191,1868,1869,1876,1883],{},[43,1870,1871,1872],{},"Markaicode — Import GGUF Models into Ollama 2026（2026-05-15）",[416,1873,1874],{"href":1874,"rel":1875},"https:\u002F\u002Fmarkaicode.com\u002Fimport-gguf-models-ollama-guide",[445],[43,1877,1878,1879],{},"ComputingForGeeks — Ollama Models Cheat Sheet 2026 ",[416,1880,1881],{"href":1881,"rel":1882},"https:\u002F\u002Fcomputingforgeeks.com\u002Follama-models-cheat-sheet",[445],[43,1884,931,1885],{},[416,1886,934],{"href":934,"rel":1887},[445],{"title":455,"searchDepth":456,"depth":456,"links":1889},[1890,1891,1892,1893,1894,1895,1896,1897,1898,1899],{"id":22,"depth":459,"text":23},{"id":38,"depth":459,"text":38},{"id":93,"depth":459,"text":93},{"id":1574,"depth":459,"text":1575},{"id":189,"depth":459,"text":189},{"id":213,"depth":459,"text":213},{"id":328,"depth":459,"text":328},{"id":363,"depth":459,"text":364},{"id":412,"depth":459,"text":412},{"id":431,"depth":459,"text":431},"\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。",[1903,1906,1909,1912],{"q":1904,"a":1905},"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":1907,"a":1908},"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":1910,"a":1911},"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":1913,"a":1914},"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。","https:\u002F\u002Fgithub.com\u002Follama\u002Follama",[477],{},[483,484,485,1919],"docker",[1921],{"plan":126,"price":971,"features":1922,"notes":1923},"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）",[981,982],{"power":489,"ux":489,"price":490,"cn_support":456,"stability":490},{"title":230,"description":1901},"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",[1930,1932,1934],{"name":1931,"url":1874,"accessed":989},"Markaicode — Import GGUF 2026",{"name":1933,"url":1881,"accessed":989},"ComputingForGeeks — Ollama Cheat Sheet 2026",{"name":993,"url":934,"accessed":989},"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",[470,1938,1939,1940,1941,502,998,1001,1460],"daemon","cli","rest-api","modelfile","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",1785660641091]