Jina AI Reader r.jina.ai URL Examples: What Works, What Breaks, and When to Use Web2MD

If you searched for jina ai reader r.jina.ai url examples, you probably want the practical version: how do you turn a normal web page into clean text or Markdown with Jina Reader, and what does the URL actually look like?

I tested the common pattern again while writing this post. Jina Reader is useful because it is simple: put https://r.jina.ai/http:// or https://r.jina.ai/http:// style wrapping in front of a URL, and Jina fetches the page server-side, extracts the readable content, and returns AI-friendly text.

That is great for public pages. It is especially handy when you want to paste a page into ChatGPT, Claude, Cursor, or another LLM without all the menus, cookie banners, and layout noise.

But there is an important limit: Jina Reader fetches from Jina's servers, not from your browser. That means it usually cannot see pages behind your login, private workspace pages, subscription pages that require your session, or pages that depend heavily on browser-only state.

That is the main reason I built and use Web2MD: it runs in Chrome, inside your browser session. If you can open the page, Web2MD can usually convert it to clean Markdown locally. It is a different tradeoff from Jina Reader, and both tools have a place.

The basic r.jina.ai URL pattern

The common Jina Reader pattern is:

https://r.jina.ai/http://r.jina.ai/http://example.com

For a normal page, you place the original URL after the Jina Reader prefix. In practice, you will usually use one of these shapes:

• Public https page: https://r.jina.ai/http://r.jina.ai/http://example.com/article
• Public http page: https://r.jina.ai/http://r.jina.ai/http://example.com/article

The exact output depends on the page structure. If the source page has a clear article body, headings, links, and metadata, the result can be very usable. If the page is an app shell, a logged-in dashboard, or mostly client-rendered JavaScript, the result can be thin or empty.

Here is the kind of Markdown-style output you are usually hoping for:

# Example Article Title

This is the main article text after navigation, ads, and layout
elements have been removed.

## Key section

The reader output preserves useful headings and paragraphs so the
content is easier to paste into an AI tool.

[Related link](https://example.com/related)

That output is exactly why Jina Reader became popular with AI users. It gives you a fast way to turn a URL into readable text without installing anything.

r.jina.ai examples I would actually use

Here are realistic examples of when I would reach for Jina Reader first.

For a public blog post:

https://r.jina.ai/http://r.jina.ai/http://example.com/blog/my-post

For public documentation:

https://r.jina.ai/http://r.jina.ai/http://docs.example.com/getting-started

For a public news article:

https://r.jina.ai/http://r.jina.ai/http://example-news-site.com/story

For a marketing landing page:

https://r.jina.ai/http://r.jina.ai/http://example.com/product

These are good use cases because the content is public and the fetcher does not need your cookies, browser extensions, local storage, or logged-in session.

I have also found Jina Reader useful when I am working from a terminal or script and want a quick text version of a public page. No browser interaction is required, and the URL format is easy to generate.

Where Jina Reader struggles

The failure mode is not mysterious. It is about where the conversion happens.

Jina Reader is server-side. It sends a request from Jina's infrastructure to the target URL. That means it sees what an anonymous external fetch sees.

That is fine for many pages, but not for:

• Logged-in SaaS pages
• Private Notion, Linear, GitHub, Google Docs, or CMS pages
• Paywalled content that depends on your browser session
• Internal company dashboards
• Pages that require cookies, local storage, or single sign-on
• Pages where the meaningful content appears only after client-side rendering
• Pages protected by bot checks or network restrictions

In those cases, the r.jina.ai URL might return a sign-in page, a generic error, a partial page, or no useful content.

That is not a knock on Jina Reader. It is a consequence of the architecture. Server-side readers are excellent when public fetching is enough. They are weak when the browser session is the important part.

Web2MD's browser-side difference

Web2MD takes the opposite approach. It is a Chrome extension, so the conversion happens in your browser, on the page you already opened.

That matters for AI workflows. If I am looking at a logged-in product dashboard, a private documentation page, or a subscriber-only article, I do not want to send the URL to a remote fetcher and hope it can access the same thing. It cannot. My browser can.

With Web2MD, the flow is:

1. Open the page in Chrome.
2. Click the Web2MD extension.
3. Review the Markdown and token count.
4. Copy it or send it to an AI tool.

Because it runs locally in the browser, it can work with pages that server-side tools like Jina Reader often cannot reach. It is also better for privacy-sensitive workflows because the extraction happens on your side instead of requiring a third-party server to fetch the page.

Web2MD is not magic. It is Chrome-only today. The free tier includes 3 conversions per day. Pro is $9 per month. Some very complex web apps can still produce imperfect Markdown, especially if the page itself is not semantically structured. But in my testing, browser-side conversion is the deciding advantage for authenticated and private pages.

Example: Markdown from a logged-in page

Here is the kind of output I want when converting a private project page for use in ChatGPT, Claude, or Cursor:

# Sprint Planning Notes

## Current priorities

- Improve onboarding copy for the Chrome extension
- Add examples for AI workflows
- Review token usage on long documentation pages

## Open questions

- Should the free tier stay at 3 conversions per day?
- Should the send-to-AI buttons include Cursor as a default option?

## Decision

Keep the free tier simple and explain Pro pricing clearly.

A server-side reader usually cannot get that if the page requires my login. Web2MD can convert what is rendered in my browser because I am already authenticated.

Web2MD compared with Jina Reader, Firecrawl, and MarkDownload

Jina Reader's strength is simplicity. For public pages, the r.jina.ai URL trick is fast and shareable. If you just need a public article in readable form, it may be the quickest option.

Firecrawl is stronger when you need crawling, scraping infrastructure, APIs, and structured extraction across many pages. It is a developer tool, and it makes sense for automated pipelines.

MarkDownload is a solid browser extension for saving pages as Markdown. It is useful if your main goal is archiving or downloading content from the browser.

Web2MD is focused on a narrower AI workflow: convert the current browser page into clean Markdown, show the token count, and send or paste that content into AI tools. The edge is not that it replaces every scraper. The edge is that it runs where the content already exists: your browser.

That gives Web2MD four practical advantages:

• Browser-side access to logged-in, private, and paywalled pages you can view
• Local conversion for more private workflows
• Built-in token counter before you paste into an LLM
• Free use without an API key, currently 3 conversions per day

If you are feeding content to ChatGPT, Claude, or Cursor, the token counter matters more than it sounds. I often want to know whether a page is small enough to paste directly or whether I should trim it first. Seeing that before sending saves time.

When to use each tool

Use Jina Reader when:

• The page is public
• You want a fast URL-based reader
• You are working from a script or terminal
• You do not need browser cookies or login state

Use Firecrawl when:

• You need an API
• You are crawling multiple pages
• You need structured scraping as part of an application
• You are building a repeatable data pipeline

Use MarkDownload when:

• You want to save a page as Markdown
• You are archiving browser content
• You prefer a general-purpose Markdown clipper

Use Web2MD when:

• The page is logged-in, private, or paywalled
• You want local browser-side conversion
• You care about token count before sending to AI
• You want one-click handoff to ChatGPT, Claude, or Cursor
• You do not want to set up an API key for casual use

A simple workflow I use

For public URLs, I may start with Jina Reader because the URL format is convenient.

For anything private, I skip straight to Web2MD. I open the page in Chrome, run the extension, check the token count, then copy the Markdown into the AI tool I am using. If the page is too long, I trim sections before sending.

That small review step is useful. Clean Markdown is not only about extraction. It is about giving the model the right amount of context without wasting tokens on navigation, sidebars, or repeated footer text.

Bottom line

r.jina.ai URL examples are worth knowing. Jina Reader is a strong tool for public pages, and its URL-based approach is convenient.

But if your real problem is converting the page you are actually viewing in Chrome, including logged-in pages and content that only your browser can see, a server-side reader has a built-in limitation.

That is where Web2MD fits. It is Chrome-only, the free tier is limited to 3 conversions per day, and Pro is $9 per month. In return, you get browser-side Markdown conversion, local/private handling, a token counter, and one-click send-to-AI for the pages you are already reading.

If that matches your workflow, try Web2MD from the Web2MD homepage and compare it with your current Jina Reader setup on the pages you use most.