离线 · 无上传 · 浏览器本地完成Offline · no uploads · runs in the browser
Markdown 转 Word（.docx）——离线、可批量、不上传Convert Markdown to Word (.docx) offline, in batches
README、需求文档、技术笔记写成 Markdown，要交给的却是一个只认 Word 的人。这条链路是 Markdown → HTML → Word：先由 marked 生成语义化 HTML，再打包成 .docx。产出的文字在 Word 里可选中、可改写、可加批注——这是它和「把 Markdown 截图成 PDF」最实际的区别。全程在你自己的电脑上完成，不发请求、不排队、不上传。A README, a spec, or a release note starts life as Markdown and ends up with someone who only reads Word. This route is Markdown → HTML → Word: marked produces semantic HTML, and the packer turns it into a .docx. The text stays selectable, editable and commentable in Word — the practical difference from photographing your Markdown into a PDF. Every step runs on your own machine: no request, no queue, no upload.
这条链路怎么走How the route runs
MarkdownMarkdown HTMLHTML Word (.docx)Word (.docx)
共 2 步，由扩展的路径搜索自动选出，中间产物不落盘。2 steps, chosen by the extension's own path search; intermediate results never touch the disk.
会保留什么What carries over
- 标题层级、有序与无序列表（含嵌套）、加粗、斜体、删除线、链接、引用块，写成 Word 的段落与样式Heading levels, ordered and unordered lists (nested included), bold, italic, strikethrough, links and block quotes become Word paragraphs and styles
- 表格转成 Word 表格；代码块与行内代码保留为等宽文字Tables become Word tables; code blocks and inline code keep a monospace face
- 以 data URI 内嵌的图片随文档一起打包，前提是类型在 `png / jpeg / gif / webp / bmp / avif / ico` 之内Images embedded as data URIs travel inside the document, provided their type is one of `png / jpeg / gif / webp / bmp / avif / ico`
- 输入 .md 按 UTF-8 → GB18030 → GBK 依次解码，老编辑器留下的 GBK 文件也能读Input is decoded UTF-8 → GB18030 → GBK, so a file saved by an older editor still reads
如实说明的限制What it does not do
- 合并单元格、脚注、mermaid 图、@mention 这类扩展语法在 HTML 一步没有对应物，会被降级成普通段落或直接丢掉Merged cells, footnotes, mermaid diagrams and @mentions have no HTML counterpart, so they are flattened into plain paragraphs or dropped
- 指向 http(s) 的外链图片在打包前被剥掉：Word 打开文件时不该替你发起网络请求，这是刻意的取舍（见 utils/converters/html-to-docx.ts 的注释）。相对路径按相对保留，图片文件要与 .docx 放在同一目录 Word 才显示得出来Remote http(s) images are stripped before packing: Word must not issue a network request while opening your file — a deliberate trade-off (see the comment in utils/converters/html-to-docx.ts). Relative paths stay relative, and the image files have to sit next to the .docx for Word to show them
- 页边距、字体、行距由 Word 的文档主题决定，CSS 里的那套版式不会跟过来Margins, fonts and line spacing come from Word’s document theme; your CSS layout does not follow
实操提示Working with it
- 一批最多 200 个文件，单个文件上限 100 MB；多个 .md 一起转就逐个产出 .docx，要一次拿全部结果可以打成一个 ZIPUp to 200 files per batch and 100 MB per file; convert a set of .md files at once and take the results as one ZIP if you prefer
- 转换结果先在工作台里预览再下载；同一份文档还能接着转成 HTML 或 PDFResults preview in the workbench before you download, and the same document can carry on to HTML or PDF
常见问题Questions people actually ask
转出来的 Word 能直接编辑吗？Is the resulting Word file editable?
能。产出是真实的段落、列表与样式，文字可选中、可改写、可加批注。若你要的是「看起来一样但不必编辑」的交付物，那一条路是 Markdown 转 PDF——但那份 PDF 是渲染出来的图像，没有文字层。Yes. The output carries real paragraphs, lists and styles, so text is selectable and commentable. If you only need something that looks the same and never gets edited, that is the Markdown to PDF route instead — but that PDF is a rendered image with no text layer.
需要装 pandoc 或 LibreOffice 吗？Do I need pandoc or LibreOffice?
不需要。整条链路是浏览器里的 JavaScript（marked 生成 HTML，再由打包器写成 .docx），没有外部进程、没有服务端、也不下载任何东西。No. The whole route is JavaScript in the browser — marked produces the HTML and the packer writes the .docx. No external process, no server, nothing fetched.
公司合同转成 Word，文件会离开本机吗？If I convert a contract, does the file leave my machine?
不会。扩展只申请 storage 权限，没有 host 权限，产物里也没有从网络取来的代码；文件读取、解析、打包都在标签页内完成。No. The extension asks for the storage permission only, has no host permissions, and ships no remotely hosted code; reading, parsing and packing all happen inside the tab.
不需要账号，也不需要联网：装好扩展后点图标，在新标签页里打开工作台，文件从磁盘直接选。No account and no network: install the extension, click its icon, and the workbench opens in a new tab reading straight off your disk.
从 Chrome 应用商店安装Install from the Chrome Web Store 安装步骤Install steps GitHub 源码Source on GitHub