20:"$Sreact.view_transition" 21:I[9754,["/_next/static/chunks/2t36f5p4pjzlb.js","/_next/static/chunks/1gyqeidjm5eoa.js","/_next/static/chunks/0_fjnxc4tt_l_.js","/_next/static/chunks/1ymt1shyhmu-4.js","/_next/static/chunks/3kv2zf0i7s2wc.js","/_next/static/chunks/1vs_3g_g75zmk.js","/_next/static/chunks/2g4dh802s8oz3.js","/_next/static/chunks/1bc3f0shukf1j.js","/_next/static/chunks/192kgs9v_t14j.js","/_next/static/chunks/0z6pw0v5c_lx5.js","/_next/static/chunks/08r60e6icmtkc.js","/_next/static/chunks/3trx6c0i7_x-3.js","/_next/static/chunks/40sro5bzp9bi_.js","/_next/static/chunks/1wgf3tk38y6cf.js"],"ProjectDetail"] 22:T6c8c, Cherry Studio · 统一调度多个前沿 LLM 的桌面工作台

UNIFIED ENTRY · 单一入口

一个桌面，调度所有前沿模型

Cherry Studio 是跨 Windows、Mac、Linux 的桌面工作台。把 OpenAI、Gemini、Anthropic 等云端大模型与 Ollama、LM Studio 本地模型接入同一个入口，一次问答即一次可追溯的路由足迹。

访问首页 GitHub 仓库

WIN / MAC / LINUX
TYPE SCRIPT
AGPL-3.0

OUTBOUND ×5 provider ROUTE FOOTPRINT · one query → many answers

DISPATCH SIGNAL / 调度层

一次提问，多路同时下发

同一个问题被分发生成到多个提供商，各自独立返回，再由工作台汇合。模型到模型之间不再来回切换标签页。

01 · 并发分发 SIMULTANEOUS CHAT

INPUT单条提示词进入入口

在统一工作区输入语句，不再为不同模型分别准备上下文。

FAN-OUT分布到多个端点并发

云端与本地模型并行接收同一条消息，互不阻塞。

CONVERGE多路回传交叉汇合

各端点独立作答后在统一视图汇合，便于对照与选择。

ENDPOINT TOPOLOGY / 端点拓扑

云端与本地，同一张图

按距离与依赖关系排布后端：需要联网的云模型与不依赖网络的本地模型，都可以作为当次调度的去向。

OpenAI Gemini Anthropic Ollama LM Studio

并联对话MULTI-MODEL

CLOUD /OpenAI、Gemini、Anthropic 等主流云服务

LOCAL /Ollama、LM Studio 本地模型

跨平台DESKTOP

OS /Windows · macOS · Linux 桌面客户端

ASSISTANT SCHEDULER / 助手调度

300+ 助手，按需唤起

预配置助手开箱即用，也能创建自定义助手；在运行期按需挂载到当次调度中，并叠加搜索与翻译等能力。

预配置助手

内置 300+ 现成助手，覆盖常见工作流，无需手动搭建即可使用。

自定义助手

按你的提示词与偏好定义专属助手，纳入统一调度清单。

全局搜索

在历史会话与本地内容间一键检索，快速回到上下文。

AI 翻译

跨语言内容直接在工作区翻译，减少在工具之间跳转。

TOOLS & MCP BUS / 工具总线

工具对接，一端即达

通过模型上下文协议让模型接入工具能力，同时直连文档与知识库处理，把调度链路扩展到内容工作流。

MCP RAG Mermaid

MCP 服务器MODEL-CONTEXT-PROTOCOL

TOOL-通过标准协议把工具能力接入模型运行期

文档处理TEXT · IMG · OFFICE · PDF

DOC /文本、图片、Office、PDF 直接处理

RAG /知识库检索增强，结合模型作答

BACKUP LEDGER / 归档台账

文件与备份，落定为账

WebDAV 文件管理与备份作为收尾台账轨，让调度走过的内容可管理、可留存。

TRACK / 台账WEBDAV

WebDAV 文件管理FILE & BACKUP

FS /通过 WebDAV 管理文件并做备份，让工作内容可留存

23:T6d5d, Cherry Studio · Desktop desk for routing multiple frontier LLMs

UNIFIED ENTRY · one entry point

One desktop,
routing every frontier model

Cherry Studio is a desktop workspace for Windows, macOS, and Linux that wires OpenAI, Gemini, and Anthropic cloud models together with local models like Ollama and LM Studio through a single entry point. Each answer is a traceable route footprint.

Visit site GitHub repo

WIN / MAC / LINUX
TYPE SCRIPT
AGPL-3.0

OUTBOUND ×5 provider ROUTE FOOTPRINT · one query → many answers

DISPATCH SIGNAL

One prompt, fanned across providers

The same message is dispatched to several providers in parallel, answered independently, then converged back in the workspace. No more tab-hopping to compare models.

01 · concurrent fan-out SIMULTANEOUS CHAT

INPUTOne prompt enters the entry point

Write once in the unified workspace instead of prepping separate context for each model.

FAN-OUTDispatched to multiple endpoints in parallel

Cloud and local models receive the same message concurrently without blocking each other.

CONVERGEReturn paths cross and combine

Each endpoint answers independently, then converges in one view for comparison and selection.

ENDPOINT TOPOLOGY

Cloud and local, on one map

Backends are arranged by distance and dependency: network-dependent cloud models and offline-capable local models can both serve as the destination of a dispatch.

OpenAI Gemini Anthropic Ollama LM Studio

Parallel chatMULTI-MODEL

CLOUD /Major cloud services incl. OpenAI, Gemini, Anthropic

LOCAL /On-device models via Ollama, LM Studio

Cross-platformDESKTOP

OS /Windows, macOS, and Linux desktop client

ASSISTANT SCHEDULER

300+ assistants, summoned on demand

Pre-configured assistants run out of the box, and you can create custom ones. Attach whichever you need to a dispatch, then layer on search and translation.

Pre-configured

A library of 300+ ready assistants covering common workflows, usable without manual setup.

Custom assistants

Define your own assistants from prompts and preferences, and schedule them like any other.

Global search

Search across history and local content to jump straight back into context.

AI translation

Translate cross-language content inside the workspace without switching tools.

TOOLS & MCP BUS

Tools wired in, at one bus

Via the Model Context Protocol, tools plug into the model runtime, and documents and knowledge bases connect directly, extending the dispatch chain into content workflows.

MCP RAG Mermaid

MCP serverMODEL-CONTEXT-PROTOCOL

TOOL-Bring tool capabilities into the model runtime through a standard protocol

Document processingTEXT · IMG · OFFICE · PDF

DOC /Handle text, images, Office files, and PDF directly

RAG /Retrieval-augmented knowledge base combined with model answers

BACKUP LEDGER

Files and backups, kept on the books

WebDAV file management and backup close out the ledger rail, so the content that flows through the dispatch chain stays manageable and preserved.

TRACKWEBDAV

WebDAV file managementFILE & BACKUP

FS /Manage files and back up via WebDAV so your work is retained

a:[["$","$20",null,{"enter":{"secondary-forward":"secondary-forward","secondary-back":"secondary-back-enter","default":"none"},"exit":{"secondary-forward":"secondary-forward","secondary-back":"secondary-back-exit","default":"none"},"update":{"secondary-forward":"secondary-forward","secondary-back":"secondary-back-update","default":"none"},"default":"none","children":["$","$L21","cmsn44bbv001m1s16pk2zx7gu:zh-CN:beautified",{"project":{"id":"cmsn44bbv001m1s16pk2zx7gu","slug":"cherryhq-cherry-studio","name":"cherry-studio","description":"支持多 LLM 的桌面 AI 客户端","authorName":"cherryhq","authorVerification":{"verified":false,"verifiedAt":null},"category":"development_tools","subcategory":"data_ai_tools","experienceUrl":"https://cherryai.com/","qualityScore":100,"featured":false,"officialSite":null,"views":{"original":{"available":true},"official":{"available":false,"reason":"disabled_by_policy"},"beautified":{"available":true,"reason":null},"defaultView":"beautified"},"beautifiedView":{"revisionId":"cmt2zgz9400bs6271en46csz1","kind":"artifact","artifact":{"schemaVersion":"localized-static-html-v4","kind":"static_html","locales":["zh-CN","en"],"defaultLocale":"zh-CN","documents":{"zh-CN":"$22","en":"$23"},"designDirections":{"schemaVersion":"dodot-design-directions-v1","directions":[{"id":"routed-dispatch","name":"The Dispatch Router","projectMetaphor":"把 Cherry Studio 视为一个统一路由中枢：每一个提示词从单一入口进入，被调度/分发到多个前沿 LLM 端点（OpenAI/Gemini/Anthropic 云端与 Ollama/LM Studio 本地），页面表达的是'一次问答产生一次可追溯的路由足迹'。","signatureVisual":"一个中心提示词节点沿辐射状'调度总线'扇形散开，连接多个后端 provider 端点；每条总线以不同曲率/夹角和信号刻度表达当次调度的去向与权重，同一提示的多路回传交叉汇合，形成一次问答的路由足迹图。","heroGeometry":"组合几何 = 居中的圆形发源点 + 从圆心放射的定向弧线簇（hub-and-spoke），弧线夹角与曲率各不相同，末端各连接一个小型端点节点；整体呈电路信号总线式的分层传播。","sectionSequence":["unified-entry-hero","dispatch-routing","providers-map","assistants-runtime","tools-mcp","backup-ledger"],"surfaceModel":"分层事务轨（layered dispatch rails）：页面横向分成若干道'路由轨'层，每层上铺排离散调度节点与刻痕分隔线，表面像电路板信号线走向，而非圆角卡片墙；节点以方形刻槽与 Mono 标签标注 provider。","palette":"工业测控冷调：墨底 #0e1116，暗轨面 #141a24，信号白 #e9eef5，路由主色琥珀信号 #f5a623，青中转 #2ec4b6，端点色用低饱和多品牌标签（#10a37f / #4285f4 / #d97757 / #0f766e / #6d28d9）仅作小芯片。","typography":"标题用紧字距的无衬线（字母间距略负、字重高），正文通栏无衬线；路由表头、provider 端点、消息头一律用等宽 mono 小字，以强化'信号/报文'语义。","desktopComposition":"左侧固定'路由控制栏'（provider 轨道翻转/多路开关），主区为横贯页面的调度扇面示意（一次问答的路由足迹），下方是端点拓扑图与助手调度清单的两栏区；再下为工具/备份台账轨。","mobileTransformation":"左侧路由控制栏折叠为顶部横向可横滚的 provider chip 条；调度扇面改为纵向'信号脉冲'逐段下发；端点拓扑与助手清单由两栏改单列；驱动改为从上到下的单条路由时间线。","mediaStrategy":"assetManifest 为空、无本地合格素材 → 移除媒体区，不加载远端图、不做占位框；以'路由足迹'派生图作为唯一标志视觉（derived_visual），不伪造产品截图。","forbiddenPatterns":["三卡指标墙（slogan + 恰好三个指标卡片）","浏览器/终端/聊天窗/编辑器壳","发光网格或粒子背景替代项目专属视觉","伪开关/搜索/输入/表单等假控件","千篇一律的圆角卡片重复堆叠"],"compositionFingerprint":"routed-dispatch/layered-rails/spoke-arcs"},{"id":"craftsmen-workbench","name":"The Craftsman's Workbench","projectMetaphor":"把 Cherry Studio 视为一张多层的工匠工作台：文档（Office/PDF/图片）、助手工具与备份抽屉像待加工的原件平铺在台面上，强调动手处理文档、WebDAV 归档与 RAG 的'生产力工作流'，而非聊天壳。","signatureVisual":"一张横向多层工作台：顶层是被切片/质检的文档原件（Office/PDF/图片，以刻痕线稿呈现），中层是一排助手工具槽，底层是 WebDAV 归档抽屉，物件以排线、折叠角等机械细节勾勒。","heroGeometry":"横向分层台面 = 数条水平'层带'（shelf 与 ledger 台账格格线叠加），物件以管线/折针对齐，底部档案抽屉；整体是一张'平面铺设'的台账而非堆叠卡片。","sectionSequence":["workbench-hero","document-processing","assistants-shelf","tools-mcp","backup-archive"],"surfaceModel":"木质/纸张台账表面：分段木纹浅底，分隔用台账格线（ledger grid）与标签条，表面带折叠、孔环、排线等持证细节；页面以'台面分层'而非圆角卡片分区。","palette":"暖工匠底 #f7f1e6，封面 #fffaf0，墨 #2b2620，工作红 #c0482f，归档棕 #8a6d3b，夹板蓝 #3f5f78，muted #8b7568。","typography":"圆角友好无衬线，标题带一点服务式字距、可用斜体强调'手工'，正文舒适行高；文档/工具标签与归档编号用 mono 小字。","desktopComposition":"顶部为工作台主视觉（多层台面带），下方文档处理走从左到右的水平管线图示，四象限工具使用区，底部归档档案排（WebDAV）。整体是横向台面叙事而非纵向营销段落。","mobileTransformation":"多层台面改纵向'抽屉'平铺（每层一个抽屉柜头、展开一块内容），文档管线转为由上而下的步骤时间线，工作台四象限在 390px 下重组为单列抽屉流。","mediaStrategy":"无本地素材 → 删除媒体区；以'工作台派遣图'派生线稿为标志视觉，文档用刻痕占位线稿表达处理关系，绝不伪造产品截图。","forbiddenPatterns":["功能卡片墙（可套用任意项目的复制）","手机/浏览器桌面壳","以渐变滤镜光效替代项目专属视觉","伪指标/伪图表/伪开关","通用网格背景替代台面语义"],"compositionFingerprint":"workbench/ledger-grid/horizontal-planes"},{"id":"orbital-constellation","name":"The Orbital Constellation","projectMetaphor":"把本地与云端模型安排进同一张轨道系统：本地模型（Ollama/LM Studio）是贴近中心的星、不依赖网络；云端模型（OpenAI/Gemini/Anthropic）是能力前沿的远星；用户围绕中心主星选择接入哪条轨道，突出 local-cloud 光谱与跨平台便携。","signatureVisual":"一张偏心星盘：中心主星（用户工作区），多个椭圆同心轨道分布 provider；近距离的本地模型与远距离的云模型列于同图，轨道交汇点用连接线标注多模型并排对话的'跨界走线'。","heroGeometry":"椭圆偏心同心轨道（offset multi-ring），主星居左略偏心，轨道用不同离心率，交汇/交叉点用点标记连接线，形成一张有主次关系的天体轨道图。","sectionSequence":["constellation-hero","local-cloud-orbit","multi-model-crosstalk","assistants-stellar","tools-mcp","backup-orbit"],"surfaceModel":"星盘/测绘图表面：深空底以极小星点与经纬测绘图格线呈现（星点只作轨道低层级辅助，不喧宾），主区为清晰轨道，表面如印刷天文图，分隔为测绘图标注而非卡片。","palette":"深空靛 #0b0f1f，墨蓝 #12172c，星点 #fff9e8，轨道主金 #f0b35c，本地青 #35d0ba，云紫 #7c6cff，muted #8a93b8。","typography":"标题用紧几何感的加粗无衬线，正文通栏无衬线；轨道刻度、provider 坐标、标签一律用 mono，呼应星图坐标语义。","desktopComposition":"主场是偏心轨道星盘主视觉，右侧垂直'轨道切换器'滑轨做 local/cloud 光谱对比；下方为多模型并排对话的跨界走线图与助手列表两栏；底部归档轨道。","mobileTransformation":"偏心星盘换为纵向'层级轨道'（每层折叠为一笔轨道，展开一块内容）；轨道切换器改顶部横滚 chip 条；并排对话的跨界走线改为单列聊天流。","mediaStrategy":"无本地素材 → 删除媒体区；以'星盘派生态'作为唯一标志视觉（derived_visual），小星点仅作轨道连接的低层级辅助，禁用远端图与占位框，不伪造产品截图。","forbiddenPatterns":["粒子化/发光网格当作替代视觉","手机/浏览器/编辑器壳","三卡指标墙","功能卡片墙","伪控制台/伪仪表板"],"compositionFingerprint":"constellation/eccentric-orbits/crosstalk-links"}],"selectedDirectionId":"routed-dispatch","selectionRationale":"在项目契合度上，Cherry Studio 的核心机制是'以统一入口接入多个前沿 LLM 并多模型同时对话'（fact: unified access to frontier LLMs + multi-model simultaneous chat），Dispatch Router 的'一个提示词被路由分发到多个端点、回传交叉汇合'最直接地表达这一机制，比 Workbench（偏向文档/工具流）与 Constellation（聚焦 local-cloud 光谱）更能承载 unified access 这一首要价值。证据契合度：Router 章节顺序覆盖统一入口、多模型路由、provider 图、300+ 助手、MCP/工具、备份等全部 verified claims，且不依赖 assetManifest（为空）而能靠'路由足迹'派生图承载标志视觉。标志视觉与 Hero 几何上，Router（中心圆 + 辐射弧 hub-and-spoke）与前代历史 seed 的'并排聊天矩阵壳'几何明显不同（后者是网格状 app shell，被本项目'浏览器/终端/聊天壳'禁用项排除）；Workbench（水平层带台账）与 Constellation（偏心椭圆轨道）也彼此几何、表面模型、手机重构均不同，三个方向非仅换色/圆角。桌面版式上 Router 以左路由控制栏 + 横贯调度扇面为骨，手机版折叠为顶部横滚 chip + 纵向信号时间线，重构逻辑独立。媒体策略三者一致地移除空媒体区并以派生图替代，但表现形制不同。相似性风险上 Router 的'分层路由轨 + 弧线'指纹与最近页面 antiTemplateContext 的空列表无冲突。综合项目机制表达力最高者选取 routed-dispatch。"},"designBrief":{"schemaVersion":"dodot-design-brief-v1","selectedDirectionId":"routed-dispatch","projectMetaphor":"把 Cherry Studio 视为一个统一路由中枢：每一个提示词从单一入口进入，被调度/分发到多个前沿 LLM 端点（OpenAI/Gemini/Anthropic 云端与 Ollama/LM Studio 本地），页面表达的是'一次问答产生一次可追溯的路由足迹'。","signatureVisual":"一个中心提示词节点沿辐射状'调度总线'扇形散开，连接多个后端 provider 端点；每条总线以不同曲率/夹角和信号刻度表达当次调度的去向与权重，同一提示的多路回传交叉汇合，形成一次问答的路由足迹图。","compositionFingerprint":"routed-dispatch/layered-rails/spoke-arcs","heroGeometry":"组合几何 = 居中的圆形发源点 + 从圆心放射的定向弧线簇（hub-and-spoke），弧线夹角与曲率各不相同，末端各连接一个小型端点节点；整体呈电路信号总线式的分层传播。","sectionSequence":["unified-entry-hero","dispatch-routing","providers-map","assistants-runtime","tools-mcp","backup-ledger"],"surfaceModel":"分层事务轨（layered dispatch rails）：页面横向分成若干道'路由轨'层，每层上铺排离散调度节点与刻痕分隔线，表面像电路板信号线走向，而非圆角卡片墙；节点以方形刻槽与 Mono 标签标注 provider。","palette":"工业测控冷调：墨底 #0e1116，暗轨面 #141a24，信号白 #e9eef5，路由主色琥珀信号 #f5a623，青中转 #2ec4b6，端点色用低饱和多品牌标签（#10a37f / #4285f4 / #d97757 / #0f766e / #6d28d9）仅作小芯片。","typography":"标题用紧字距的无衬线（字母间距略负、字重高），正文通栏无衬线；路由表头、provider 端点、消息头一律用等宽 mono 小字，以强化'信号/报文'语义。","desktopComposition":"左侧固定'路由控制栏'（provider 轨道翻转/多路开关），主区为横贯页面的调度扇面示意（一次问答的路由足迹），下方是端点拓扑图与助手调度清单的两栏区；再下为工具/备份台账轨。","mobileTransformation":"左侧路由控制栏折叠为顶部横向可横滚的 provider chip 条；调度扇面改为纵向'信号脉冲'逐段下发；端点拓扑与助手清单由两栏改单列；驱动改为从上到下的单条路由时间线。","mediaStrategy":"assetManifest 为空、无本地合格素材 → 移除媒体区，不加载远端图、不做占位框；以'路由足迹'派生图作为唯一标志视觉（derived_visual），不伪造产品截图。","evidencePlan":["unified-entry-hero：来自 summary 的核心声明 '支持多 LLM 的统一桌面客户端 / 跨 Windows·Mac·Linux'（跨平台 + unified access）。","dispatch-routing / providers-map：多模型同时对话 + 支持 OpenAI/Gemini/Anthropic 云端与 Ollama/LM Studio 本地，各 provider 以 Mono 芯片标注（对应 verified claim）。","assistants-runtime：300+ 预配置助手 + 自定义助手 + AI 翻译/全局搜索，取自 readme 与 verified claims。","tools-mcp：MCP 服务器、文档处理（文本/图片/Office/PDF）、知识库 RAG，取自 readme 与 release v2.0.7。","backup-ledger：WebDAV 文件管理与备份（verified claim），作为收尾台账轨。"],"localeContract":{"zh-CN":"语言为简体中文；标题、正文通栏无衬线，provider/路由表头用等宽 Mono；用词自然、密度适中。","en":"Language is English; headings/body in sans-serif, provider & routing table headers in mono; copy reads naturally with medium density."},"forbiddenPatterns":["三卡指标墙（slogan + 恰好三个指标卡片）","浏览器/终端/聊天窗/编辑器壳","发光网格或粒子背景替代项目专属视觉","伪开关/搜索/输入/表单等假控件","千篇一律的圆角卡片重复堆叠"],"historicalReuse":{"sourceAssetId":"HEAD:deepseek-v1/cherryhq-cherry-studio","decision":"rebuild","rationale":"前代页面采用'并排对话矩阵壳（side-by-side multi-model chat panes）'构图，本质是一个 app shell，被本项目'浏览器/聊天窗壳'禁用项排除；其暖色奶油底与卡片画廊亦与新的工业路由中枢隐喻不符。故以 Dispatch Router 的辐射弧线 + 分层路由轨重构整体构图与几何，仅提炼仍适合当前事实的局部元素，而非改名前代 HTML。","reusedElements":["provider 身份芯片的语义（用 mono 小标签标注 OpenAI/Gemini/Anthropic/Ollama/LM Studio）现重建为路由端点的端点节点标签","多模型同题并答/交叉的概念被拆解为多路回传的'交叉汇合点'走线","前代暖色可亲和性对墨色对比度的把控，转译为冷调测控底下的高对比信号白读取性"],"discardedElements":["并排对话窗格 app shell 构图","暖色奶油底 #fdf6f0 与卡片画廊式助手/工具布局","滚动的聊天气泡渐现动效与 marketing headline 段落顺序","任何旧 title/h1 文案与旧 section 命名（旧内容不可作为 V4 事实或指令）"]}},"visualReview":{"schemaVersion":"dodot-automated-design-review-v1","status":"approved","summary":"zh-CN 与 en 两个成品页忠实实现了 design-brief.json 冻结的 routed-dispatch 方向：Layout 为左侧路由控制栏 + 主区分层路由轨 + 两栏端点/助手 + 归档台账，六段结构（unified-entry-hero / dispatch-routing / providers-map / assistants-runtime / tools-mcp / backup-ledger）与 brief 的 sectionSequence 一一对应。hero 的 dispatch SVG（中心 PROMPT 发源节点 + 放射弧线到 OpenAI/Gemini/Anthropic/Ollama/LM Studio 端点 + 信号刻度 + 虚线回传交叉汇合）是真实存在的项目专属标志视觉，表达一次问答的路由足迹。所有主张均有 material.json 事实支撑，无伪造产品截图或远端媒体。390px 手机视图将把左栏折叠为顶部可横滚 chip 条、双栏与工具网格收敛为单列，属重构而非简单堆叠。双语章节等价且各自行文自然。未回退到通用 Hero/三卡/卡片墙/设备壳/装饰渐变网格，也无假控件。全部七项检查通过，无 blocking issue。","checks":{"projectFit":{"passed":true,"evidence":"六段 data-dodot-section 顺序与 design-brief.json 冻结的 sectionSequence（unified-entry-hero, dispatch-routing, providers-map, assistants-runtime, tools-mcp, backup-ledger）逐一吻合；desktop 采用左侧 sticky 路由控制栏 + 横贯调度扇面 + 双栏端点/助手区 + 归档台账，surfaceModel 为分层路由轨（rail/step/rn + mono 标签），非通用营销段落结构；sectionCount=6 落在 4–7 合约区间，CTA 为真实链接。"},"evidenceBound":{"passed":true,"evidence":"跨 Windows/macOS/Linux（material summary 跨平台）、统一接入 OpenAI/Gemini/Anthropic 云端 + Ollama/LM Studio 本地（facts）、300+ 助手与自定义助手（fact）、全局搜索与 AI 翻译（fact）、MCP 与文档处理（文本/图片/Office/PDF）与知识库 RAG 与 Mermaid（readme/release 事实）、WebDAV 文件管理与备份（fact）、AGPL-3.0 与 TypeScript（fact）均可在 material.json 中找到出处；dispatch/拓扑/SVG 均为 derived_visual 示意，未伪造产品截图；assetManifest 为空故未引入媒体区，外链仅指向材料内的 cherryai.com 与 GitHub 仓库；qa 报告 networkRequests 为空、无 broken image。"},"signatureVisual":{"passed":true,"evidence":"hero 的 dispatch SVG（viewBox 0 0 880 360）含中心 PROMPT 发源节点、三条曲率/夹角不同的放射弧线分别连到 OpenAI/Gemini/Anthropic/Ollama/LM Studio 端点盒、弧上信号刻度圆点、虚线回传交叉汇合路径，配 ROUTE FOOTPRINT 说明与 OUTBOUND x5 provider 图注，直接实现 brief 的 hub-and-spoke 调度足迹视觉，两语言文件均有等价 SVG 与 aria-label。"},"mobileTransformation":{"passed":true,"evidence":"@media (max-width:820px) 将 .routerail 隐藏并显示顶部可横滚 provider chip 条（.chipstrip，flex non-wrap + overflow-x auto + 隐藏滚动条），.two 与 .tools 收敛为单列，dispatch 扇面随容器缩放，fan-caption 折列；QA 报告 mobile 390 视口 scrollWidth=390 无横向溢出，重构决策（左栏→顶部 chip 条、双栏→单列）与 brief 的 mobileTransformation 一致，并非桌面列简单堆叠。"},"localeParity":{"passed":true,"evidence":"zh-CN（lang=zh-CN）与 en（lang=en）共享同一 brief 与六段章节序、同一 CSS/结构/拓扑图/标志视觉；中文文案通顺自然、英文文案为等价独立写法且密度适中（zh textLength 1920 vs en 3683），provider/端点/图标等 mono 标记两语言一致；sourceCode aria-label 与页脚声明均本地化。"},"antiTemplate":{"passed":true,"evidence":"无 slogan+三卡指标墙；无浏览器/终端/聊天气泡/编辑器壳（调度扇面为项目专属 derived 视觉且具 aria-label，是示意非壳）；无功能卡片墙（为 rail/step 台账轨且 radius 2px 刻蚀感）；背景 44px 网格透明度极低（rgba .035）仅作支撑层，主视觉为调度足迹而非装饰渐变/粒子替代；mono provider/端点标签为语义，无伪图表。antiTemplateContext 空列表，指纹 routed-dispatch/layered-rails/spoke-arcs 未冲突；historal 决策 rebuild 剔除前代聊天窗壳，符合 skill 禁止项。"},"noFakeUi":{"passed":true,"evidence":"provider 芯片（.chip、.chipstrip、.rail-pulse）均为 aria-hidden 的装饰 span 而非按钮/开关/输入；路由轨链接为真实 # 锚点，CTA 为真实外链；无伪搜索/切换/仪表盘/客户 logo/测试评论；dispatch 与拓扑 SVG 为带 aria-label 的信息图；QA 报告两语言各视口 fakeControlCount=0、emptyLinkCount=0、consoleErrors=[]，无假控件。"}},"issues":[]},"repairEvidence":[{"attempt":1,"before":{"qaPassed":false,"designReviewApproved":false,"qaEvidence":{"zh-CN":[{"viewport":"desktop","passed":true,"localePassed":true,"metrics":{"language":"zh-CN","h1Count":1,"h2Count":5,"sectionCount":6,"textLength":1920,"cjkCount":660,"latinLetterCount":873,"scrollWidth":1280,"clientWidth":1280,"scrollHeight":3896,"h1Top":176.59375,"bottomInset":314.703125,"brokenImageCount":0,"missingAltCount":0,"fakeControlCount":0,"emptyLinkCount":0},"screenshotHashes":{"top":"62a6022217fa6a7b0a6eb2bb87257b36cf0641cdd6ae4f152a1eeee358326258","middle":"c5b8be761ca45bcdf46fc86bd60555741d50ad883e3b9d6f643eaa8f77665e52","bottom":"f70fa165804fca056c24fccc71733fa6c32751cb9e864e0d1d449ec5cea4e7b8"},"consoleErrors":[],"networkRequests":[]},{"viewport":"mobile","passed":false,"localePassed":true,"metrics":{"language":"zh-CN","h1Count":1,"h2Count":5,"sectionCount":6,"textLength":1762,"cjkCount":660,"latinLetterCount":744,"scrollWidth":469,"clientWidth":390,"scrollHeight":4653,"h1Top":172.59375,"bottomInset":374.984375,"brokenImageCount":0,"missingAltCount":0,"fakeControlCount":0,"emptyLinkCount":0},"screenshotHashes":{"top":"b69a021a6fa3e26acee303f6d17b073f4d4820f7988343712668bfe78a93e9f1","middle":"9b275c96216322c93ace9b33b33facbfaad96bf0f1df7c82bf0bb4c25353022f","bottom":"1fe3acc295da1bcd189e78999d5e8cc11ae37812ad5b89cc258b0c71efcca940"},"consoleErrors":[],"networkRequests":[]}],"en":[{"viewport":"desktop","passed":false,"localePassed":true,"metrics":{"language":"en","h1Count":1,"h2Count":5,"sectionCount":6,"textLength":3683,"cjkCount":0,"latinLetterCount":2977,"scrollWidth":1516,"clientWidth":1280,"scrollHeight":4047,"h1Top":176.59375,"bottomInset":314.203125,"brokenImageCount":0,"missingAltCount":0,"fakeControlCount":0,"emptyLinkCount":0},"screenshotHashes":{"top":"bdbfb169855aeb90bfbd2d5ee77334f7ef6ef7723c67685367a4c45e7f03a66f","middle":"c52b1992289f35c02ab40061b53e5d5e8a65a0ce547d1c74cd9ae342465c96bb","bottom":"35980b32e3e22567ef1e99a32e9c5d9ed7801fc5b6d05a9647af6a23698d547e"},"consoleErrors":[],"networkRequests":[]},{"viewport":"mobile","passed":false,"localePassed":true,"metrics":{"language":"en","h1Count":1,"h2Count":5,"sectionCount":6,"textLength":3526,"cjkCount":0,"latinLetterCount":2849,"scrollWidth":692,"clientWidth":390,"scrollHeight":5091,"h1Top":172.59375,"bottomInset":411.078125,"brokenImageCount":0,"missingAltCount":0,"fakeControlCount":0,"emptyLinkCount":0},"screenshotHashes":{"top":"f57103b294fbe44e11a91e053163ff937c74a405dd75995b958625618aca069d","middle":"402c00e83c381dd78ce5a9eb165b9c9e5ff082c0f8d2be270b0e586f59797c6f","bottom":"9bd26a479ee2f34785ff56c5b69eaa341433ef038a6859f113d0bce29cdf063f"},"consoleErrors":[],"networkRequests":[]}]},"visualReview":null},"after":{"qaPassed":true,"designReviewApproved":true,"qaEvidence":{"zh-CN":[{"viewport":"desktop","passed":true,"localePassed":true,"metrics":{"language":"zh-CN","h1Count":1,"h2Count":5,"sectionCount":6,"textLength":1920,"cjkCount":660,"latinLetterCount":873,"scrollWidth":1280,"clientWidth":1280,"scrollHeight":3928,"h1Top":176.59375,"bottomInset":314.75,"brokenImageCount":0,"missingAltCount":0,"fakeControlCount":0,"emptyLinkCount":0},"screenshotHashes":{"top":"62a6022217fa6a7b0a6eb2bb87257b36cf0641cdd6ae4f152a1eeee358326258","middle":"756fc217ac76c7814d6ae17498e443b0a3f94fbe587d3193329543595f50b7d1","bottom":"daeb8603f273619dc48db1c5cdbfdb7ab4f87e06f9cbcbab3c8240b5213e82a8"},"consoleErrors":[],"networkRequests":[]},{"viewport":"mobile","passed":true,"localePassed":true,"metrics":{"language":"zh-CN","h1Count":1,"h2Count":5,"sectionCount":6,"textLength":1762,"cjkCount":660,"latinLetterCount":744,"scrollWidth":390,"clientWidth":390,"scrollHeight":4736,"h1Top":172.59375,"bottomInset":374.203125,"brokenImageCount":0,"missingAltCount":0,"fakeControlCount":0,"emptyLinkCount":0},"screenshotHashes":{"top":"780a18e397e1e75d107318f88d3bdaeff75c7b8cf38599ade9a72d6bbc26fc61","middle":"dadf7f8bd4eee91e69739e6453d0829da842ff7774ba23d192362ad8bd24322c","bottom":"b6e39d6615023fd8d73edb18bf3bb2815d8097e27570e12ac6b86dfc0b99ce76"},"consoleErrors":[],"networkRequests":[]}],"en":[{"viewport":"desktop","passed":true,"localePassed":true,"metrics":{"language":"en","h1Count":1,"h2Count":5,"sectionCount":6,"textLength":3683,"cjkCount":0,"latinLetterCount":2977,"scrollWidth":1280,"clientWidth":1280,"scrollHeight":4105,"h1Top":176.59375,"bottomInset":314.65625,"brokenImageCount":0,"missingAltCount":0,"fakeControlCount":0,"emptyLinkCount":0},"screenshotHashes":{"top":"bdbfb169855aeb90bfbd2d5ee77334f7ef6ef7723c67685367a4c45e7f03a66f","middle":"cc82de31191b58eba69d06019887d83f7e34d29366f9673b9b3ebdb73d5a6b60","bottom":"302d0a63748070b50eb0c306822f600349f4bf10fc7fc8082848839d95d024a7"},"consoleErrors":[],"networkRequests":[]},{"viewport":"mobile","passed":true,"localePassed":true,"metrics":{"language":"en","h1Count":1,"h2Count":5,"sectionCount":6,"textLength":3526,"cjkCount":0,"latinLetterCount":2849,"scrollWidth":390,"clientWidth":390,"scrollHeight":5306,"h1Top":172.59375,"bottomInset":411.515625,"brokenImageCount":0,"missingAltCount":0,"fakeControlCount":0,"emptyLinkCount":0},"screenshotHashes":{"top":"56a64b97e7b70a52aa031784e7c7a8d2e7772e446c3ba56e796a359a819ca749","middle":"9f4df90e9bae5264ff438215482647f243817698f9a52e1e5ca57ef0d26c4bef","bottom":"9273704cdeef75e6b4526c99ce02998661d5d8b3eed5d62b205fe021d5ed9855"},"consoleErrors":[],"networkRequests":[]}]},"visualReview":{"schemaVersion":"dodot-automated-design-review-v1","status":"approved","summary":"zh-CN 与 en 两个成品页忠实实现了 design-brief.json 冻结的 routed-dispatch 方向：Layout 为左侧路由控制栏 + 主区分层路由轨 + 两栏端点/助手 + 归档台账，六段结构（unified-entry-hero / dispatch-routing / providers-map / assistants-runtime / tools-mcp / backup-ledger）与 brief 的 sectionSequence 一一对应。hero 的 dispatch SVG（中心 PROMPT 发源节点 + 放射弧线到 OpenAI/Gemini/Anthropic/Ollama/LM Studio 端点 + 信号刻度 + 虚线回传交叉汇合）是真实存在的项目专属标志视觉，表达一次问答的路由足迹。所有主张均有 material.json 事实支撑，无伪造产品截图或远端媒体。390px 手机视图将把左栏折叠为顶部可横滚 chip 条、双栏与工具网格收敛为单列，属重构而非简单堆叠。双语章节等价且各自行文自然。未回退到通用 Hero/三卡/卡片墙/设备壳/装饰渐变网格，也无假控件。全部七项检查通过，无 blocking issue。","checks":{"projectFit":{"passed":true,"evidence":"六段 data-dodot-section 顺序与 design-brief.json 冻结的 sectionSequence（unified-entry-hero, dispatch-routing, providers-map, assistants-runtime, tools-mcp, backup-ledger）逐一吻合；desktop 采用左侧 sticky 路由控制栏 + 横贯调度扇面 + 双栏端点/助手区 + 归档台账，surfaceModel 为分层路由轨（rail/step/rn + mono 标签），非通用营销段落结构；sectionCount=6 落在 4–7 合约区间，CTA 为真实链接。"},"evidenceBound":{"passed":true,"evidence":"跨 Windows/macOS/Linux（material summary 跨平台）、统一接入 OpenAI/Gemini/Anthropic 云端 + Ollama/LM Studio 本地（facts）、300+ 助手与自定义助手（fact）、全局搜索与 AI 翻译（fact）、MCP 与文档处理（文本/图片/Office/PDF）与知识库 RAG 与 Mermaid（readme/release 事实）、WebDAV 文件管理与备份（fact）、AGPL-3.0 与 TypeScript（fact）均可在 material.json 中找到出处；dispatch/拓扑/SVG 均为 derived_visual 示意，未伪造产品截图；assetManifest 为空故未引入媒体区，外链仅指向材料内的 cherryai.com 与 GitHub 仓库；qa 报告 networkRequests 为空、无 broken image。"},"signatureVisual":{"passed":true,"evidence":"hero 的 dispatch SVG（viewBox 0 0 880 360）含中心 PROMPT 发源节点、三条曲率/夹角不同的放射弧线分别连到 OpenAI/Gemini/Anthropic/Ollama/LM Studio 端点盒、弧上信号刻度圆点、虚线回传交叉汇合路径，配 ROUTE FOOTPRINT 说明与 OUTBOUND x5 provider 图注，直接实现 brief 的 hub-and-spoke 调度足迹视觉，两语言文件均有等价 SVG 与 aria-label。"},"mobileTransformation":{"passed":true,"evidence":"@media (max-width:820px) 将 .routerail 隐藏并显示顶部可横滚 provider chip 条（.chipstrip，flex non-wrap + overflow-x auto + 隐藏滚动条），.two 与 .tools 收敛为单列，dispatch 扇面随容器缩放，fan-caption 折列；QA 报告 mobile 390 视口 scrollWidth=390 无横向溢出，重构决策（左栏→顶部 chip 条、双栏→单列）与 brief 的 mobileTransformation 一致，并非桌面列简单堆叠。"},"localeParity":{"passed":true,"evidence":"zh-CN（lang=zh-CN）与 en（lang=en）共享同一 brief 与六段章节序、同一 CSS/结构/拓扑图/标志视觉；中文文案通顺自然、英文文案为等价独立写法且密度适中（zh textLength 1920 vs en 3683），provider/端点/图标等 mono 标记两语言一致；sourceCode aria-label 与页脚声明均本地化。"},"antiTemplate":{"passed":true,"evidence":"无 slogan+三卡指标墙；无浏览器/终端/聊天气泡/编辑器壳（调度扇面为项目专属 derived 视觉且具 aria-label，是示意非壳）；无功能卡片墙（为 rail/step 台账轨且 radius 2px 刻蚀感）；背景 44px 网格透明度极低（rgba .035）仅作支撑层，主视觉为调度足迹而非装饰渐变/粒子替代；mono provider/端点标签为语义，无伪图表。antiTemplateContext 空列表，指纹 routed-dispatch/layered-rails/spoke-arcs 未冲突；historal 决策 rebuild 剔除前代聊天窗壳，符合 skill 禁止项。"},"noFakeUi":{"passed":true,"evidence":"provider 芯片（.chip、.chipstrip、.rail-pulse）均为 aria-hidden 的装饰 span 而非按钮/开关/输入；路由轨链接为真实 # 锚点，CTA 为真实外链；无伪搜索/切换/仪表盘/客户 logo/测试评论；dispatch 与拓扑 SVG 为带 aria-label 的信息图；QA 报告两语言各视口 fakeControlCount=0、emptyLinkCount=0、consoleErrors=[]，无假控件。"}},"issues":[]}},"changedFiles":["artifact/zh-CN.html","artifact/en.html"]}],"provenance":{"executor":"opencode","model":"deepseek/deepseek-v4-flash","opencodeVersion":"1.1.58","promptVersion":"dodot-showcase-deepseek-v4","inputHash":"e7100a8b736e1164bfbe2a689142f57fab33d24271dfb82bb8d8f9c46228d8e0","artifactHash":"453bc47d7ff29c83af092eaa1703c45c82335d3120bdb0c860fa18f56fbdec1d","designSkill":"dodot-showcase-design","historicalSeed":{"source":"git_head","sourceCanonicalKey":"github:cherryhq/cherry-studio","assetId":"HEAD:deepseek-v1/cherryhq-cherry-studio","sourceSchemaVersion":"v1","sourceGeneratorVersion":"anthropic:claude-fable-5","sourceArtifactHash":"da38a700ea842e9561135df8184533dd70656b824d938db43a083b985733e1a9","disposition":"repair_or_regenerate","seedPayloadHash":"2cb05311e74470e77975e822c1fea52470897a96f04d9b40861c3b6fd28d16b2"}},"qaEvidence":{"zh-CN":[{"viewport":"desktop","passed":true,"localePassed":true,"metrics":{"language":"zh-CN","h1Count":1,"h2Count":5,"sectionCount":6,"textLength":1920,"cjkCount":660,"latinLetterCount":873,"scrollWidth":1280,"clientWidth":1280,"scrollHeight":3928,"h1Top":176.59375,"bottomInset":314.75,"brokenImageCount":0,"missingAltCount":0,"fakeControlCount":0,"emptyLinkCount":0},"screenshotHashes":{"top":"62a6022217fa6a7b0a6eb2bb87257b36cf0641cdd6ae4f152a1eeee358326258","middle":"756fc217ac76c7814d6ae17498e443b0a3f94fbe587d3193329543595f50b7d1","bottom":"daeb8603f273619dc48db1c5cdbfdb7ab4f87e06f9cbcbab3c8240b5213e82a8"},"consoleErrors":[],"networkRequests":[]},{"viewport":"mobile","passed":true,"localePassed":true,"metrics":{"language":"zh-CN","h1Count":1,"h2Count":5,"sectionCount":6,"textLength":1762,"cjkCount":660,"latinLetterCount":744,"scrollWidth":390,"clientWidth":390,"scrollHeight":4736,"h1Top":172.59375,"bottomInset":374.203125,"brokenImageCount":0,"missingAltCount":0,"fakeControlCount":0,"emptyLinkCount":0},"screenshotHashes":{"top":"780a18e397e1e75d107318f88d3bdaeff75c7b8cf38599ade9a72d6bbc26fc61","middle":"dadf7f8bd4eee91e69739e6453d0829da842ff7774ba23d192362ad8bd24322c","bottom":"b6e39d6615023fd8d73edb18bf3bb2815d8097e27570e12ac6b86dfc0b99ce76"},"consoleErrors":[],"networkRequests":[]}],"en":[{"viewport":"desktop","passed":true,"localePassed":true,"metrics":{"language":"en","h1Count":1,"h2Count":5,"sectionCount":6,"textLength":3683,"cjkCount":0,"latinLetterCount":2977,"scrollWidth":1280,"clientWidth":1280,"scrollHeight":4105,"h1Top":176.59375,"bottomInset":314.65625,"brokenImageCount":0,"missingAltCount":0,"fakeControlCount":0,"emptyLinkCount":0},"screenshotHashes":{"top":"bdbfb169855aeb90bfbd2d5ee77334f7ef6ef7723c67685367a4c45e7f03a66f","middle":"cc82de31191b58eba69d06019887d83f7e34d29366f9673b9b3ebdb73d5a6b60","bottom":"302d0a63748070b50eb0c306822f600349f4bf10fc7fc8082848839d95d024a7"},"consoleErrors":[],"networkRequests":[]},{"viewport":"mobile","passed":true,"localePassed":true,"metrics":{"language":"en","h1Count":1,"h2Count":5,"sectionCount":6,"textLength":3526,"cjkCount":0,"latinLetterCount":2849,"scrollWidth":390,"clientWidth":390,"scrollHeight":5306,"h1Top":172.59375,"bottomInset":411.515625,"brokenImageCount":0,"missingAltCount":0,"fakeControlCount":0,"emptyLinkCount":0},"screenshotHashes":{"top":"56a64b97e7b70a52aa031784e7c7a8d2e7772e446c3ba56e796a359a819ca749","middle":"9f4df90e9bae5264ff438215482647f243817698f9a52e1e5ca57ef0d26c4bef","bottom":"9273704cdeef75e6b4526c99ce02998661d5d8b3eed5d62b205fe021d5ed9855"},"consoleErrors":[],"networkRequests":[]}]}},"generatorVersion":"opencode-1.1.58:deepseek/deepseek-v4-flash:dodot-showcase-deepseek-v4","publishedAt":"2026-08-20T12:34:27.242Z"},"rawExtraction":{"projectRevisionId":"cmt2zgz8c00bm6271j3jomfdc","projectVersion":1,"sources":[{"canonicalKey":"github:cherryhq/cherry-studio","role":"primary","revisionVersion":1,"url":"https://github.com/CherryHQ/cherry-studio"}],"facts":[{"key":"repository_owner","value":"cherryhq","evidence":[{"path":"repository.owner","excerpt":"cherryhq"}]},{"key":"repository_name","value":"cherry-studio","evidence":[{"path":"repository.name","excerpt":"cherry-studio"}]},{"key":"description","value":"AI productivity studio with smart chat, autonomous agents, and 300+ assistants. Unified access to frontier LLMs","evidence":[{"path":"repository.description","excerpt":"AI productivity studio with smart chat, autonomous agents, and 300+ assistants. Unified access to frontier LLMs"}]},{"key":"primary_language","value":"TypeScript","evidence":[{"path":"repository.language","excerpt":"TypeScript"}]},{"key":"license","value":"AGPL-3.0","evidence":[{"path":"repository.license","excerpt":"AGPL-3.0"}]},{"key":"topics","value":"agent-skills, ai-agent, claude-code, codex, deepseek, deepseek-harness, hermes-agent, skills, vibe-coding","evidence":[{"path":"repository.topics","excerpt":"agent-skills, ai-agent, claude-code, codex, deepseek, deepseek-harness, hermes-agent, skills, vibe-coding"}]},{"key":"source_count","value":"1","evidence":[{"path":"projectRevision.sources","excerpt":"github:cherryhq/cherry-studio"}]}],"taxonomy":{"category":"development_tools","subcategory":"data_ai_tools","confidence":0.65,"evidence":[{"path":"repository.description","excerpt":"AI productivity studio with smart chat, autonomous agents, and 300+ assistants. Unified access to frontier LLMs"}]},"lifecycle":{"stage":"active","confidence":0.75,"evidence":[{"path":"repository.pushedAt","excerpt":"2026-08-17T12:50:28.000Z"}]},"userIntents":[{"slug":"build_software","label":"构建或改进软件","confidence":0.65,"evidence":[{"path":"repository.description","excerpt":"AI productivity studio with smart chat, autonomous agents, and 300+ assistants. Unified access to frontier LLMs"}]},{"slug":"automate_work","label":"自动化工作流程","confidence":0.65,"evidence":[{"path":"repository.description","excerpt":"AI productivity studio with smart chat, autonomous agents, and 300+ assistants. Unified access to frontier LLMs"}]},{"slug":"create_content","label":"创作内容或设计","confidence":0.65,"evidence":[{"path":"repository.description","excerpt":"AI productivity studio with smart chat, autonomous agents, and 300+ assistants. Unified access to frontier LLMs"}]},{"slug":"analyze_data","label":"分析与理解数据","confidence":0.65,"evidence":[{"path":"repository.description","excerpt":"AI productivity studio with smart chat, autonomous agents, and 300+ assistants. Unified access to frontier LLMs"}]},{"slug":"learn_explore","label":"学习与探索","confidence":0.65,"evidence":[{"path":"repository.description","excerpt":"AI productivity studio with smart chat, autonomous agents, and 300+ assistants. Unified access to frontier LLMs"}]}],"readme":{"headings":["🍒 Cherry Studio","🌠 Screenshot","🌟 Key Features","📝 Roadmap","🌈 Theme","🤝 Contributing","Getting Started","🔧 Developer Co-creation Program","Contributor Rewards Program","Growing Together & Future Plans","How to Get Started?","🏢 Enterprise Edition","Core Advantages","✨ Online Demo","Version Comparison","Get the Enterprise Edition","🔗 Related Projects","🚀 Contributors","📊 GitHub Stats","⭐️ Star History"],"commands":[]},"changeSet":[{"field":"authorName","beforeHash":"74234e98afe7498fb5daf1f36ac2d78acc339464f950703b8c019892f982b90b","afterHash":"9d9d62b2d7558369a763f4762979deee325d7848f464d690a96f361a2cfff20f"},{"field":"description","beforeHash":"74234e98afe7498fb5daf1f36ac2d78acc339464f950703b8c019892f982b90b","afterHash":"ce396ef7c0ae592540bcaa2fd7db0b1a8608ade722dc23b74f50aaa095e44c53"},{"field":"experienceUrl","beforeHash":"74234e98afe7498fb5daf1f36ac2d78acc339464f950703b8c019892f982b90b","afterHash":"f96542a999cb009e900a720f7c17f6d1a576b0f3902b3fe899116f190cd91390"},{"field":"name","beforeHash":"74234e98afe7498fb5daf1f36ac2d78acc339464f950703b8c019892f982b90b","afterHash":"f709a2d3d478445b84e5cf5d6e85ff6bcb202be73565ef70215d00b6c73822d3"},{"field":"repository","beforeHash":"74234e98afe7498fb5daf1f36ac2d78acc339464f950703b8c019892f982b90b","afterHash":"29b116d0228d79047b8eca9a9622d56772fe2cd3b1accdecb3838784f33ecb0f"},{"field":"sourceCount","beforeHash":"74234e98afe7498fb5daf1f36ac2d78acc339464f950703b8c019892f982b90b","afterHash":"6b86b273ff34fce19d6b804eff5a3f5747ada4eaa22f1d49c01e52ddb7875b4b"}]},"publishedAt":"2026-08-10T10:55:52.371Z","updatedAt":"2026-08-21T13:26:48.983Z","summary":{"id":"cmt2zgz8u00bo6271wpoxfmwh","projectType":"product","oneLine":"支持多 LLM 的桌面 AI 客户端","overview":"Cherry Studio 是一款跨平台桌面客户端，聚合 OpenAI、Gemini、Anthropic 等主流云模型及 Ollama 本地模型。内置 300+ 预配置助手、文档处理、MCP 服务器、知识库与翻译工具，适合日常对话、内容创作和轻量开发辅助。","highlights":["300+ 预配置助手","多模型同时对话","支持本地模型","内置 MCP 服务器"],"bestFor":["需要统一入口使用多种 LLM 的用户","希望本地运行模型的开发者","需要文档处理与知识库的办公场景"],"caveats":["部分高级功能（如企业版）需付费","移动端与插件系统仍在路线图中","模型能力取决于所接入的提供商"],"categorySuggestion":"productivity","subcategorySuggestion":"automation_workflows","lifecycleStages":["idea_requirements","product_design","frontend","backend_data"],"audienceIntents":["discover","evaluate","build","integrate"],"confidence":0.85,"evidence":[{"claim":"Cherry Studio 是一款支持多 LLM 提供商的桌面客户端，可在 Windows、Mac 和 Linux 上使用。","sourceType":"readme","locator":"https://github.com/CherryHQ/cherry-studio"},{"claim":"提供 300+ 预配置 AI 助手，支持自定义助手创建和多模型同时对话。","sourceType":"readme","locator":"https://github.com/CherryHQ/cherry-studio"},{"claim":"支持文档处理（文本、图片、Office、PDF）、WebDAV 文件管理和备份、Mermaid 图表可视化。","sourceType":"readme","locator":"https://github.com/CherryHQ/cherry-studio"},{"claim":"集成 MCP（模型上下文协议）服务器、全局搜索、主题管理和 AI 翻译等实用工具。","sourceType":"readme","locator":"https://github.com/CherryHQ/cherry-studio"},{"claim":"支持本地模型（Ollama、LM Studio）以及 OpenAI、Gemini、Anthropic 等主流云服务。","sourceType":"readme","locator":"https://github.com/CherryHQ/cherry-studio"},{"claim":"最新版本 v2.0.7 包含代理运行时、知识库 RAG、代码伴侣（集成 DeepSeek Harness）等更新。","sourceType":"release","locator":"https://github.com/CherryHQ/cherry-studio/releases/tag/v2.0.7"}],"publishedAt":"2026-08-17T13:57:34.806Z"},"github":{"url":"https://github.com/CherryHQ/cherry-studio","ownerAvatarUrl":"https://avatars.githubusercontent.com/u/187777663?v=4","starsCount":50640,"forksCount":4803,"language":"TypeScript","syncedAt":"2026-08-17T13:54:09.508Z","license":"AGPL-3.0","defaultBranch":"main","archived":false,"disabled":false,"lastChangedAt":"2026-08-17T12:50:28.000Z","missing":false},"tags":[{"name":"Vibe coding","slug":"vibe-coding","verification":"verified","evidenceUrl":"https://github.com/CherryHQ/cherry-studio","evidenceExcerpt":"GitHub topic: vibe-coding"}],"updates":[{"id":"cmsnc7wj2007m1s16ntzg3h0l","kind":"head_updated","title":"默认分支有新提交","summary":null,"sourceUrl":"https://github.com/CherryHQ/cherry-studio/commit/e883e0db27325fbea3d9a644ea35e6cd368497b3","occurredAt":"2026-08-10T14:36:49.000Z"}]},"entryPoint":"ranking","claim":{"state":"unclaimed","verified":false},"initialView":"beautified","community":{"viewer":null,"state":{"likesCount":0,"creativeCount":0,"followersCount":0,"commentsCount":0,"liked":false,"creative":false,"followed":false},"comments":{"comments":[],"nextCursor":null}}}]}],"$L24"]