https://r.jina.ai/http://www.example.com: What It Does and When to Use a Browser Markdown Converter Instead
https://r.jina.ai/http://www.example.com: What It Does and When to Use a Browser Markdown Converter Instead
If you have seen a URL like https://r.jina.ai/http://www.example.com, you have probably found one of the simplest ways to turn a public web page into Markdown.
The pattern is easy:
- Take a normal page URL, like
http://www.example.com - Put
https://r.jina.ai/in front of it - Open the new URL
- Copy the Markdown-like text into ChatGPT, Claude, Cursor, or another AI tool
I tested this flow with the classic Example Domain page, and it works exactly as expected. The page is public, tiny, and has no login wall, so Jina Reader can fetch it from its own servers and return a clean text version.
For a page like Example Domain, that is hard to beat for speed.
But the same pattern has limits. It depends on a remote service being able to fetch the page. That means it usually cannot see the same page you see after logging in, opening a paid article, loading a private dashboard, or using an internal app. That is where a browser-side converter like Web2MD is useful.
Web2MD is a Chrome extension that converts the page already open in your browser into clean Markdown for AI tools. In my testing, the practical difference is not just formatting. It is access.
What https://r.jina.ai/http://www.example.com returns
When I opened https://r.jina.ai/http://www.example.com, the output was a compact Markdown version of the Example Domain page.
A simplified version looks like this:
# Example Domain
This domain is for use in illustrative examples in documents. You may use this domain in literature without prior coordination or asking for permission.
[More information...](https://www.iana.org/domains/example)
That is exactly the kind of output people want for AI prompting: title, body text, and links, without the visual layout noise.
For public pages, Jina Reader is a strong tool. It is fast, requires no browser extension, and is easy to automate. If your use case is converting public documentation, blog posts, or simple pages from scripts, Jina Reader is often a good fit.
The key phrase is "public pages."
The important limit: Jina sees the web, not your browser
The URL https://r.jina.ai/http://www.example.com asks Jina Reader to fetch http://www.example.com from Jina's servers.
That is fine for Example Domain. It is not the same as converting the page you are looking at in Chrome.
Here are common cases where that difference matters:
- A logged-in SaaS dashboard
- A private Notion page
- A course lesson behind an account
- A paid research article
- A GitHub issue in a private repository
- An internal company wiki
- A customer support ticket
- A page that only renders after client-side JavaScript
- A page customized for your session
In those cases, a server-side reader may get a login page, an error, a bot check, or a generic public version. It does not have your browser session, your cookies, or the exact rendered DOM you are seeing.
Web2MD works differently. It runs in Chrome and converts the current page locally in your browser. If you can open the page, Web2MD can usually convert the visible content into Markdown.
That is the main reason I reach for Web2MD instead of a remote URL reader when I am preparing context for ChatGPT, Claude, or Cursor from real work pages.
Browser-side conversion is not just about access
Access is the obvious advantage, but privacy is the second one.
With a URL like https://r.jina.ai/http://www.example.com, the page URL is sent to a third-party service, and the service fetches the content. That may be acceptable for public pages. It may not be acceptable for private work.
Web2MD keeps the conversion in your browser. You are not sending the source page to a scraping API just to get Markdown. For sensitive pages, that is a simpler mental model: convert locally, review the output, then decide what to paste into an AI tool.
This matters because web-to-Markdown is often used right before asking an AI model to analyze something. A local conversion step gives you one more chance to remove private details before pasting.
What Web2MD output looks like
I tested Web2MD on simple pages and heavier pages. The goal is not to preserve every pixel. The goal is to produce clean Markdown that an AI model can understand without wasting tokens on navigation, ads, duplicate menus, and layout clutter.
For a simple page, the output is similar to what you would expect:
# Example Domain
This domain is for use in illustrative examples in documents. You may use this domain in literature without prior coordination or asking for permission.
[More information...](https://www.iana.org/domains/example)
On longer pages, Web2MD is more useful because it includes a built-in token counter. That sounds small until you are trying to fit a long article, docs page, or support thread into an AI prompt.
A real workflow looks like this:
# Product Changelog
## July updates
We improved export speed, added keyboard shortcuts, and fixed several rendering issues in tables.
## Known limitations
Some embedded widgets are skipped because they are loaded in isolated frames.
## Next steps
Use the feedback form in the dashboard to report pages that do not convert cleanly.
Before sending that to an AI tool, I can see the approximate token count and decide whether to paste the whole thing, trim sections, or summarize first.
That is one of Web2MD's practical advantages over plain copy and paste. It does not just give you Markdown. It helps you manage the context window.
Comparing Jina Reader, Firecrawl, MarkDownload, and Web2MD
There is no single best tool for every web-to-Markdown job.
Jina Reader is excellent when you want a quick Markdown view of a public URL. The https://r.jina.ai/http://www.example.com style is memorable and easy to share. It is also useful for automation and quick testing.
Firecrawl is strong for developer workflows. It can crawl sites, extract structured data, and support more advanced scraping pipelines. If you are building a backend system that ingests public websites at scale, Firecrawl is worth looking at.
MarkDownload is a useful browser extension for saving pages as Markdown. It is especially handy if your main goal is archiving pages into a note-taking system.
Web2MD is focused on a slightly different job: getting the current browser page into clean Markdown for AI tools. Its edge is browser-side conversion, privacy, a built-in token counter, and one-click send-to-AI. It also has a free tier with 3 conversions per day and does not require an API key.
The tradeoffs are real. Web2MD is Chrome-only today. The free tier is limited to 3 conversions per day. Pro is $9 per month. If you need server-side crawling or automated batch extraction, Web2MD is not trying to replace Firecrawl. If you only need a public page once, Jina Reader may be faster.
But if the page is private, logged in, paywalled, personalized, or just easier to access in your browser, Web2MD is the better fit.
When I would use each tool
Use Jina Reader when:
- The page is public
- You want a fast URL-based conversion
- You do not need browser session access
- You are comfortable sending the URL to a remote reader
- You want something simple like
https://r.jina.ai/http://www.example.com
Use Firecrawl when:
- You are building a scraping or ingestion pipeline
- You need crawling, extraction, or automation
- You are working with public sites at scale
- You are comfortable using an API-based workflow
Use MarkDownload when:
- You want to save pages as Markdown
- You are building a personal archive
- You are clipping pages into notes
- AI prompting is not the main workflow
Use Web2MD when:
- The page is already open in Chrome
- The page requires login or paid access
- You care about local browser-side conversion
- You want to check token count before prompting
- You want to send page context to ChatGPT, Claude, or Cursor quickly
Why this matters for AI prompts
AI tools work better with clean input. Raw web pages include navigation, cookie banners, sidebars, scripts, repeated links, and layout text. Copying from the browser often brings weird spacing or misses important headings.
Markdown gives the model structure:
- Headings show hierarchy
- Lists stay readable
- Links keep their labels and destinations
- Tables are easier to inspect
- Long pages are easier to trim
That structure is useful whether you are asking ChatGPT to summarize a page, Claude to compare documents, or Cursor to turn documentation into code changes.
The token counter matters because context is limited. Even large-context models have practical limits. A 20,000-token dump of a page may be slower, more expensive, and less accurate than a 4,000-token Markdown version that keeps the main content.
A practical recommendation
If your target is a public page like Example Domain, try the Jina Reader pattern first. https://r.jina.ai/http://www.example.com is simple, fast, and good enough for many public pages.
If the page is private, logged in, paywalled, or something you would not want to send through a server-side reader, use a browser-side tool. That is the gap Web2MD is built for.
It is not magic, and it is not unlimited. It is Chrome-only, the free plan gives you 3 conversions per day, and Pro is $9 per month. Some complex pages with embedded apps or isolated frames may still need cleanup. But for the everyday job of turning the page in front of you into AI-ready Markdown, it is the workflow I prefer.
If you want to try it, visit Web2MD, install the Chrome extension, and convert a page you already use with ChatGPT, Claude, or Cursor. Start with the free tier and see whether the local Markdown plus token counter fits your workflow.