Open the current-page tool and choose MD
Open a blog post, news article or tutorial in Chrome or Edge. Click the WeChat to Markdown extension icon and select “Open current page tools.” When body detection succeeds, a tool panel appears on the page. Click “MD” to export Markdown. To save illustrations locally, first enable “Download article images (MD / HTML).”
Ordinary webpages and single WeChat articles share the same free export allowance. See the plan details for the allowance and advanced features. Ordinary pages are read after your explicit action by default. The optional “Automatically show on this site” setting applies to the site you choose and is not required for an export.
Turn a readable webpage into a local Markdown file
- Choose an article with a complete body
Install the extension from the installation page. Open a blog post, news detail page or tutorial, rather than search results, a category page or a comment list. Complete any login or expansion steps you already have permission to use, and confirm the content you need has loaded.
- Open the current-page tool yourself
Click the extension icon in the toolbar and select “Open current page tools.” The extension detects article content and removes distractions such as navigation. If it cannot find enough body text, it may show a limited panel. The presence of a panel alone does not prove that the article was fully detected.
- Choose your image preference and export MD
Click “MD” directly to keep online image references. For offline illustrations, enable “Download article images (MD / HTML)” before exporting. This produces a ZIP containing the body and successfully downloaded images.
- Verify before organizing
Open the exported file and compare its title, key paragraphs, code and images with the webpage. To filter and organize articles later, you can also click “Add to local article library,” then use “Open article library.”
Example: save a Java tutorial with code
Suppose you are reading “Getting started with Java scheduled tasks.” The page has a title, two heading levels, a code block, an execution diagram and a sidebar of recommended posts. The goal is to keep the tutorial body without copying the sidebar and navigation.
# Getting started with Java scheduled tasks
## Create a scheduled task
```java
System.out.println("task started");
```
## Check the result
These are the verification steps from the tutorial.This illustrates the intended content structure; it is not a guarantee for every site. With image downloads off, images still use online URLs. When downloads are enabled and succeed, the Markdown in the ZIP references local images under _assets/stable-article-id/. Title, author and date extraction depends on the information the webpage provides.
Check four things after exporting
- Body boundaries: the article begins in the right place, its ending is complete, and large amounts of navigation or recommended content have not been mixed into it.
- Meaning and structure: headings, lists, quotations and links still convey the original meaning. Check indentation, line breaks and special characters in code, and inspect complex tables row by row.
- Source information: the source link points to this article, and the author and publish date match the page. Add missing information yourself when needed; do not treat absent metadata as verified.
- Images: if you chose an image ZIP, extract it fully and check the files and relative paths. Disconnect and reopen the preview to test offline access. Failed image downloads may still reference online URLs.
Which webpages may not convert completely?
No panel appears, or the page is unsupported
Confirm that the current page is an ordinary http:// or https:// webpage, then select “Open current page tools” again. Browser internal pages, extension stores, extension pages and local file:// files are restricted by browser permissions and are outside this workflow.
The export is very short or cuts off the article
The tool identifies an article from content already loaded on the page. Very short text, embedded frames, content loaded only after scrolling, paginated articles and interactive content may not be captured completely. Inspect what the browser can display, then trigger the tool again. An incomplete body should not count as a successful archive.
The layout differs from the original webpage
Markdown focuses on content structure. Multiple columns, complex styles, interactive charts and media players are not guaranteed to survive. For HTML or print-style output, see the HTML format guide and PDF format guide. Changing formats does not bypass the page’s loading requirements or access restrictions.
The extension does not bypass logins, paywalls or permission checks, and it does not promise support for every site. Conversion cannot recover an unreadable or deleted article that was never archived. For a reproducible detection failure, send the page address, expected body area and actual result through the support page.
Keep the next copy where you will use it
For simple reference storage, keep the Markdown and image archive. To add it to a personal knowledge base, follow the Obsidian directory guide or SiYuan import workflow. WeChat articles have a dedicated adapter; start with the WeChat-to-Markdown guide for those pages.
This guide describes the official store product. The open-source Community MVP is not the store’s free plan. See About and your installed version for the ordinary-webpage feature boundary. How to judge a fit, and the paid-article limit, are in how to choose a tool.