Where converted Markdown goes next

Which use case fits you

Pick by the object you need next week, not by the app you have open today. If the object is a claim you can check, you want a source note and a chat that had to quote it. That is AI research. If the object is a Markdown file inside a folder Obsidian already watches, you want Obsidian. If the object is a diff, a fixture, or a script that fetches one public URL, you want Developers.

The three overlap, and the overlap is sequential. A developer reading a public RFC can follow the research steps, store the file with the vault steps, and later replace the manual download with an API call. Starting with the API because it sounds more automated is the wrong order when you have not yet seen what the extractor does to that particular docs theme.

None of these pages is a crawler, a vault sync, or a chat product. WebInk converts a page you can already reach. The guides cover the conversion, the send, the HTML comparison, and the save. Come back to a guide when a use case says “see the save guide” rather than repeating every filename rule here.

What each page teaches

AI research is a one-source habit. You convert a public link with URL to Markdown, or you convert the open tab with the extension when the site required a login. You put the title and the URL into the file with a template, you read the preview for leftover sidebars, and you keep the file after the model answers. The page is specific about the extension’s stock prompts: Summarize, Extract Key Info, and Mind Map substitute the URL, while Translate to English and Fix Formatting substitute the Markdown body. It also covers ZIP export when a figure has to remain readable after the host changes the image URL.

Obsidian is a filing habit. The Default template is the article body alone. Basic Frontmatter adds a YAML block with title, url, and publishedTime, and that title is inserted without extra quotes, so a colon in the title can break the YAML until you edit the template. Download .zip produces title.md beside an images folder, with successful images rewritten to relative links. Obsidian renders those standard Markdown images. WebInk does not emit ![[wikilinks]], and it does not choose a vault folder. The page says when Obsidian Web Clipper is the better capture.

Developers is a pipeline habit for a single page. The HTML tool converts a fragment in the browser. The URL tool and GET /api/markdown fetch a public URL on the server. POST /api/markdown can also take a template string or transform options, with a real caveat: sending both means the transform step replaces the templated Markdown. The extension remains the path for a rendered tab, including pages behind your login. The page tells you to pin one converter if a later diff must mean the docs changed rather than the tool changed.

Quick start for each destination

Research note, public article. Paste the URL into the URL tool. Select or create a template only if you are in the extension. On the web tool the download is the article Markdown the API returned. Add the source URL at the top of the file if it is not already there. Ask the model questions that require a quotation from a heading you can see in the file. Keep the file in a sources/ folder named by date and site.

Vault note, with figures. Convert in the extension after the images have loaded. Switch the template to Basic Frontmatter if you want properties your vault can search, and quote the title in the template when it contains a colon. Choose Download .zip. Unzip so the Markdown file and the images folder stay siblings, then move that pair into the vault. Open the note and confirm one image renders before you archive the zip.

Docs snippet for a pull request. If you can copy the HTML, use HTML to Markdown and download the .md. If you have a public URL, use the URL tool or the API and read the fences before you commit. If the docs site only renders after login, use the extension and do not send that text to a third-party chat unless the page is allowed to leave your machine.

A short file from the research path looks like this once a template has added the source. The body is illustrative.

---
title: "Cache invalidation notes"
url: https://example.com/posts/cache-invalidation
publishedTime: "2026-04-02"
---

## What changed

Writers delete the key on update. A short TTL only hides a stale value until that delete exists.

The built-in Basic Frontmatter template does not add those quotes for you. The sample shows the safer shape after you edit the template. The Default template would have omitted the YAML block entirely.

Common mistakes

• Saving fifty URLs as one chore. Each conversion is one page. Batch the work by repeating a single-page checklist, not by pointing WebInk at a sitemap.
• Sending the stock Summarize prompt and treating the answer as a reading of the article. Put {{markdown}} in the prompt when the body has to be in the chat. Details are in the send guide.
• Unzipping into the vault root so images/ collides with another note’s images. Give each capture its own folder.
• Hashing API output and extension output in the same fixture folder. They are related products and different converters.
• Assuming a leading H1 will always be in an extension download. The extension converter removes a leading # title line from the body. Put {{title}} in the template when the file must name itself.
• Using the public URL tool on a page you would not want the server to fetch. Your session stays in the browser. The extension is the path that reads that session’s document.

Limits that apply to all three

WebInk does not log into sites for you, click through a consent wall, or follow a “next chapter” pager. Scroll a lazy page before you convert if the missing part loads on scroll. Images that fail to download during ZIP export keep their original URL, and the extension toast reports how many failed. Data-URL images stay inline. A Notion export control is not currently available.

Conversion, preview, copy, templates, the table of contents, and Send to AI are described on the features pages. Plan information is on Pricing. If you are new to the product, the webpage to Markdown guide is still the right first read, and the Chrome Web Store listing is where the extension is installed.

FAQ